{ "api_version": "1", "updated": "2026-10-29", "latest_event": "2026-10-28", "rights": "© Frontier Log, all rights reserved.", "benchmark": { "slug": "livecodebench-no-release", "name": "LiveCodeBench, no release named", "family": "livecodebench", "unit": "%", "higher_is_better": true, "status": "retired", "active": false, "superseded_by": null, "retired": "2026-05-20", "benchmark_url": "https://livecodebench.github.io", "introduced": "2024-03-12", "creator": "UC Berkeley, MIT and Cornell", "introduced_source": "https://arxiv.org/abs/2403.07974", "settings": [ { "setting": "custom-window", "name": "Lab-chosen window", "url": "https://frontierlog.ai/benchmark/livecodebench-no-release", "results": 29 }, { "setting": "window-unstated", "name": "Window not stated", "url": "https://frontierlog.ai/benchmark/livecodebench-no-release/window-unstated", "results": 22 } ] }, "total": 51, "results": [ { "id": "gemma-2-27b:livecodebench-no-release:window-unstated:1", "model": "gemma-2-27b", "model_name": "Gemma 2 27B", "model_date": "2024-06-27", "event": "https://frontierlog.ai/event/2024-06-27-gemma-2", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 20.4, "unit": "%", "printed": "20.4", "reported_by": "google", "run_by": null, "label": "Gemma 2 27B", "source": "https://storage.googleapis.com/deepmind-media/gemma/Gemma3Report.pdf", "source_title": "Gemma 3 Technical Report", "source_date": "2025-03-12", "where": "Gemma 3 technical report (PDF), Table 6 (instruction-tuned models), row 'LiveCodeBench', column 'Gemma 2 27B'", "own": false, "primary": true, "parallel": false, "notes": "instruction-tuned; 0-shot; average of 8 samples", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-20", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_capture": "https://frontierlog.ai/captures/f86f02802b4c2394.png" }, { "id": "gemma-2-27b:livecodebench-no-release:window-unstated:2", "model": "gemma-2-27b", "model_name": "Gemma 2 27B", "model_date": "2024-06-27", "event": "https://frontierlog.ai/event/2024-06-27-gemma-2", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 29.0, "unit": "%", "printed": "29.0", "reported_by": "google", "run_by": null, "label": "Gemma 2 27B", "source": "https://storage.googleapis.com/deepmind-media/gemma/Gemma3Report.pdf", "source_title": "Gemma 3 Technical Report", "source_date": "2025-03-12", "where": "Gemma 3 Technical Report, Table 18 (instruction-tuned models, more internal and external benchmarks), page 23, row 'LiveCodeBench', column 'Gemma 2 27B'", "own": false, "primary": false, "parallel": false, "notes": "chain of thought; instruction-tuned; 0-shot; average of 8 samples", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-29", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [] }, { "id": "llama-3-1-405b:livecodebench-no-release:custom-window:1", "model": "llama-3-1-405b", "model_name": "Llama 3.1 405B", "model_date": "2024-07-23", "event": "https://frontierlog.ai/event/2024-07-23-llama-3-1-405b", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 27.7, "unit": "%", "printed": "27.7", "reported_by": "meta", "run_by": null, "label": "Llama 3.1 405B", "source": "https://ai.meta.com/blog/llama-4-multimodal-intelligence/", "source_title": "The Llama 4 herd: The beginning of a new era of natively multimodal AI innovation", "source_date": "2025-04-05", "where": "'Llama 4 Scout instruction-tuned benchmarks' table (image), row 'LiveCodeBench (10/01/2024-02/01/2025)', column 'Llama 3.1 405B'", "own": false, "primary": true, "parallel": false, "notes": "problems from Oct 2024 to Feb 2025; 0-shot; averaged over multiple generations", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-20", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/47ea0cbe5c73836d.png", "source_image_original": "https://scontent-ord5-1.xx.fbcdn.net/v/t39.2365-6/488658055_1347378876402143_3412007366291908454_n.png?_nc_cat=109&ccb=1-7&_nc_sid=e280be&_nc_ohc=OBma4RowFAYQ7kNvwEDKyuH&_nc_oc=AdqPMOh8lXBK3lg7quc_HNoP-49jOvB8JczkL0IsoavEpo4Igwy0r1vrenrHwo1uFmA&_nc_zt=14&_nc_ht=scontent-ord5-1.xx&_nc_gid=pzE_LW4H_OG--eyEc4cgFA&_nc_ss=7f289&oh=00_AQNipOGf-EGDwZJ5T-zywEffyFQYUnzut_O2CKWGuWBPNg&oe=6AF0271F" }, { "id": "gemini-2-0-flash:livecodebench-no-release:custom-window:1", "model": "gemini-2-0-flash", "model_name": "Gemini 2.0 Flash", "model_date": "2024-12-11", "event": "https://frontierlog.ai/event/2024-12-11-gemini-2-0-flash", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 35.1, "unit": "%", "printed": "35.1%", "reported_by": "google", "run_by": null, "label": "Gemini 2.0 Flash Experimental", "source": "https://blog.google/innovation-and-ai/models-and-research/google-deepmind/google-gemini-ai-update-december-2024/", "source_title": "Introducing Gemini 2.0: our new AI model for the agentic era", "source_date": "2024-12-11", "where": "benchmark table (image), row 'Code, LiveCodeBench (Code Generation)', column 'Gemini 2.0 Flash Experimental'", "own": true, "primary": true, "parallel": false, "notes": "experimental version; code generation subset; problems from 06/01/2024 to 10/05/2024", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-29", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/9a121b3d0bb030e8.gif", "source_image_original": "https://storage.googleapis.com/gweb-uniblog-publish-prod/original_images/gemini_benchmarks_narrow_light2x.gif" }, { "id": "deepseek-v3:livecodebench-no-release:window-unstated:1", "model": "deepseek-v3", "model_name": "DeepSeek-V3", "model_date": "2024-12-26", "event": "https://frontierlog.ai/event/2024-12-26-deepseek-v3", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 40.5, "unit": "%", "printed": "40.5", "reported_by": "deepseek", "run_by": null, "label": "DeepSeek-V3", "source": "https://api-docs.deepseek.com/news/news1226/", "source_title": "Introducing DeepSeek-V3", "source_date": "2024-12-26", "where": "benchmark table (image), row 'LiveCodeBench(Pass@1-COT)', column 'DeepSeek-V3'", "own": true, "primary": true, "parallel": false, "notes": "chain-of-thought; 8K output limit", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-29", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [ { "url": "https://github.com/deepseek-ai/DeepSeek-V3", "name": "DeepSeek-V3 card", "title": "deepseek-ai/DeepSeek-V3 GitHub repository", "lab": "deepseek", "date": "2024-12-26", "printed": "40.5", "where": "README Chat model evaluation table, column \"DeepSeek V3\", row \"LiveCodeBench (Pass@1-COT)\"" } ], "source_image": "https://frontierlog.ai/sources/f55e6180f36f8bcf.jpeg", "source_image_original": "https://cdn.deepseek.com/api-docs/ds_v3_benchmark_table_en.jpeg" }, { "id": "deepseek-v3:livecodebench-no-release:window-unstated:2", "model": "deepseek-v3", "model_name": "DeepSeek-V3", "model_date": "2024-12-26", "event": "https://frontierlog.ai/event/2024-12-26-deepseek-v3", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 37.6, "unit": "%", "printed": "37.6", "reported_by": "deepseek", "run_by": null, "label": "DeepSeek-V3", "source": "https://api-docs.deepseek.com/news/news1226/", "source_title": "Introducing DeepSeek-V3", "source_date": "2024-12-26", "where": "benchmark table (image), row 'LiveCodeBench (Pass@1)', column 'DeepSeek-V3'", "own": true, "primary": false, "parallel": false, "notes": "without chain-of-thought; 8K output limit", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-29", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [ { "url": "https://github.com/deepseek-ai/DeepSeek-V3", "name": "DeepSeek-V3 card", "title": "deepseek-ai/DeepSeek-V3 GitHub repository", "lab": "deepseek", "date": "2024-12-26", "printed": "37.6", "where": "README Chat model evaluation table, column \"DeepSeek V3\", row \"LiveCodeBench (Pass@1)\"" } ], "source_image": "https://frontierlog.ai/sources/f55e6180f36f8bcf.jpeg", "source_image_original": "https://cdn.deepseek.com/api-docs/ds_v3_benchmark_table_en.jpeg" }, { "id": "codestral-25-01:livecodebench-no-release:window-unstated:1", "model": "codestral-25-01", "model_name": "Codestral 25.01", "model_date": "2025-01-13", "event": "https://frontierlog.ai/event/2025-01-13-codestral-25-01", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 37.9, "unit": "%", "printed": "37.9%", "reported_by": "mistral", "run_by": null, "label": "Codestral-2501", "source": "https://mistral.ai/news/codestral-2501/", "source_title": "Codestral 25.01", "source_date": "2025-01-13", "where": "'Overview' benchmark table (HTML), Python group, row 'Codestral-2501', column 'LiveCodeBench'", "own": true, "primary": true, "parallel": false, "notes": "", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "double_entry", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_capture": "https://frontierlog.ai/captures/dee44180c5a2a639.png" }, { "id": "deepseek-r1:livecodebench-no-release:custom-window:1", "model": "deepseek-r1", "model_name": "DeepSeek-R1", "model_date": "2025-01-20", "event": "https://frontierlog.ai/event/2025-01-20-deepseek-r1", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 65.9, "unit": "%", "printed": "65.9", "reported_by": "deepseek", "run_by": null, "label": "DeepSeek R1", "source": "https://github.com/deepseek-ai/DeepSeek-R1/blob/main/DeepSeek_R1.pdf", "source_title": "DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning", "source_date": "2025-01-20", "where": "technical report, Table 4, row 'LiveCodeBench (Pass@1-COT)', column 'DeepSeek R1'", "own": true, "primary": true, "parallel": false, "notes": "problems from Aug 2024 to Jan 2025; chain-of-thought; 32K output limit", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://github.com/deepseek-ai/DeepSeek-R1", "name": "DeepSeek-R1 card", "title": "DeepSeek-R1 GitHub repository", "lab": "deepseek", "date": "2025-01-20", "printed": "65.9", "where": "README Evaluation table, row \"LiveCodeBench (Pass@1-COT)\", column \"DeepSeek R1\"" } ] }, { "id": "qwen2-5-14b-instruct-1m:livecodebench-no-release:custom-window:1", "model": "qwen2-5-14b-instruct-1m", "model_name": "Qwen2.5-14B-Instruct-1M", "model_date": "2025-01-27", "event": "https://frontierlog.ai/event/2025-01-27-qwen2-5-1m", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 38.6, "unit": "%", "printed": "38.6", "reported_by": "alibaba", "run_by": null, "label": "Qwen2.5-14B-1M", "source": "https://qwen.ai/blog?id=qwen2.5-1m", "source_title": "Qwen2.5-1M: Deploy Your Own Qwen with Context Length up to 1M Tokens", "source_date": "2025-01-27", "where": "short-context benchmark table image (short_result), Coding Tasks block, row 'LiveCodeBench', column 'Qwen2.5-14B-1M'", "own": true, "primary": true, "parallel": false, "notes": "problems from May 2023 to Sep 2024", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-28", "check": "human", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen2.5-1M/Qwen2_5_1M_Technical_Report.pdf", "name": "Qwen2.5-1M card", "title": "Qwen2.5-1M Technical Report", "lab": "alibaba", "date": "2025-01-27", "printed": "38.6", "where": "Qwen2.5-1M Technical Report, Table 6, row LiveCodeBench, column Qwen2.5-14B-1M; text says LiveCodeBench 2305-2409" } ], "source_image": "https://frontierlog.ai/sources/06383364ff31282b.png", "source_image_original": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen2.5-1M/short_result.png" }, { "id": "qwen2-5-max:livecodebench-no-release:custom-window:1", "model": "qwen2-5-max", "model_name": "Qwen2.5-Max", "model_date": "2025-01-28", "event": "https://frontierlog.ai/event/2025-01-28-qwen2-5-max", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 38.7, "unit": "%", "printed": "38.7", "reported_by": "alibaba", "run_by": null, "label": "Qwen2.5-Max", "source": "https://qwen.ai/blog?id=qwen2.5-max", "source_title": "Qwen2.5-Max: Exploring the Intelligence of Large-scale MoE Model", "source_date": "2025-01-28", "where": "instruct-model bar chart image, group 'LiveCodeBench (24.08-24.11)', first bar 'Qwen2.5-Max' (red), value label", "own": true, "primary": true, "parallel": false, "notes": "instruct model; problems from Aug to Nov 2024", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-19", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/98f8f3360dba3e6a.jpg", "source_image_original": "https://qianwen-res.oss-accelerate-overseas.aliyuncs.com/Qwen2.5-max-instruct.jpg" }, { "id": "qwq-32b:livecodebench-no-release:custom-window:1", "model": "qwq-32b", "model_name": "QwQ-32B", "model_date": "2025-03-06", "event": "https://frontierlog.ai/event/2025-03-06-qwq-32b", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 63.4, "unit": "%", "printed": "63.4", "reported_by": "alibaba", "run_by": null, "label": "QwQ-32B", "source": "https://qwen.ai/blog?id=qwq-32b", "source_title": "QwQ-32B: Embracing the Power of Reinforcement Learning", "source_date": "2025-03-06", "where": "bar chart image under 'Performance', group 'LiveCodeBench (24.08-25.02)', first bar 'QwQ-32B' (red), value label", "own": true, "primary": true, "parallel": false, "notes": "problems from Aug 2024 to Feb 2025", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-19", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/415acc7917a65f6b.jpg", "source_image_original": "https://qianwen-res.oss-accelerate-overseas.aliyuncs.com/qwq-32b-final.jpg" }, { "id": "gemma-3-27b:livecodebench-no-release:window-unstated:1", "model": "gemma-3-27b", "model_name": "Gemma 3 27B", "model_date": "2025-03-12", "event": "https://frontierlog.ai/event/2025-03-12-gemma-3", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 29.7, "unit": "%", "printed": "29.7", "reported_by": "google", "run_by": null, "label": "Gemma 3 27B", "source": "https://storage.googleapis.com/deepmind-media/gemma/Gemma3Report.pdf", "source_title": "Gemma 3 Technical Report", "source_date": "2025-03-12", "where": "Table 6 (instruction-tuned models, zero-shot benchmarks), page 6, row 'LiveCodeBench', column 'Gemma 3 27B'", "own": true, "primary": true, "parallel": false, "notes": "instruction-tuned; 0-shot; average of 8 samples", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-29", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [] }, { "id": "gemma-3-27b:livecodebench-no-release:window-unstated:2", "model": "gemma-3-27b", "model_name": "Gemma 3 27B", "model_date": "2025-03-12", "event": "https://frontierlog.ai/event/2025-03-12-gemma-3", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 39.0, "unit": "%", "printed": "39.0", "reported_by": "google", "run_by": null, "label": "Gemma 3 27B", "source": "https://storage.googleapis.com/deepmind-media/gemma/Gemma3Report.pdf", "source_title": "Gemma 3 Technical Report", "source_date": "2025-03-12", "where": "Table 18 (instruction-tuned models, more internal and external benchmarks), page 23, row 'LiveCodeBench', column 'Gemma 3 27B'", "own": true, "primary": false, "parallel": false, "notes": "chain of thought; instruction-tuned; 0-shot; average of 8 samples", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-29", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [] }, { "id": "llama-nemotron-super-49b-v1:livecodebench-no-release:window-unstated:1", "model": "llama-nemotron-super-49b-v1", "model_name": "Llama Nemotron Super 49B v1", "model_date": "2025-03-18", "event": "https://frontierlog.ai/event/2025-03-18-llama-nemotron", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 45.5, "unit": "%", "printed": "45.5", "reported_by": "nvidia", "run_by": null, "label": "Llama Nemotron Super 49B v1", "source": "https://developer.nvidia.com/blog/build-more-accurate-and-efficient-ai-agents-with-the-new-nvidia-llama-nemotron-super-v1-5/", "source_title": "Build More Accurate and Efficient AI Agents with the New NVIDIA Llama Nemotron Super v1.5", "source_date": "2025-07-29", "where": "Figure 2 bar chart (image), value label of the 'Llama Nemotron Super 49B v1' bar (lightest, rightmost in group) in group 'LiveCodeBench'", "own": false, "primary": true, "parallel": false, "notes": "", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-19", "corroborated_by": [ { "url": "https://huggingface.co/nvidia/Llama-3_3-Nemotron-Super-49B-v1_5/blob/420ba7d28211abf116b8b103ab700d92619daf98/README.md", "name": "Llama Nemotron Super v1.5 card", "title": "nvidia/Llama-3_3-Nemotron-Super-49B-v1_5 model card (revision of 2025-10-15)", "lab": "nvidia", "date": "2025-07-29", "printed": "45.5", "where": "accuracy_chart.png bar chart, LiveCodeBench group, 'Llama Nemotron Super 49B v1' bar label" }, { "url": "https://huggingface.co/nvidia/Llama-3_3-Nemotron-Super-49B-v1_5-FP8/blob/04822723e77e036ddf2d24e83c6d469d3b009252/README.md", "name": "Llama Nemotron Super v1.5 card", "title": "nvidia/Llama-3_3-Nemotron-Super-49B-v1_5-FP8 model card (revision of 2025-10-15)", "lab": "nvidia", "date": "2025-07-29", "printed": "45.5", "where": "accuracy_chart.png bar chart, LiveCodeBench group, 'Llama Nemotron Super 49B v1' bar label" }, { "url": "https://huggingface.co/nvidia/Llama-3_3-Nemotron-Super-49B-v1_5-NVFP4/blob/f0caaafd5152e07527c7a14e04aa67823107529f/README.md", "name": "Llama Nemotron Super v1.5 card", "title": "nvidia/Llama-3_3-Nemotron-Super-49B-v1_5-NVFP4 model card (revision of 2025-11-27)", "lab": "nvidia", "date": "2025-07-29", "printed": "45.5", "where": "accuracy_chart.png bar chart, LiveCodeBench group, 'Llama Nemotron Super 49B v1' bar label" } ], "source_image": "https://frontierlog.ai/sources/d431180dea13f63f.png", "source_image_original": "https://developer-blogs.nvidia.com/wp-content/uploads/2025/07/image-29.png" }, { "id": "deepseek-v3-0324:livecodebench-no-release:window-unstated:1", "model": "deepseek-v3-0324", "model_name": "DeepSeek-V3-0324", "model_date": "2025-03-25", "event": "https://frontierlog.ai/event/2025-03-25-deepseek-v3-0324", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 49.2, "unit": "%", "printed": "49.2", "reported_by": "deepseek", "run_by": null, "label": "DeepSeek-V3-0324", "source": "https://api-docs.deepseek.com/news/news250325/", "source_title": "DeepSeek-V3-0324 Release", "source_date": "2025-03-25", "where": "benchmark bar chart (image), group 'LiveCodeBench (Pass@1)', value label on the 'DeepSeek-V3-0324' bar", "own": true, "primary": true, "parallel": false, "notes": "", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-29", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [ { "url": "https://huggingface.co/deepseek-ai/DeepSeek-V3-0324", "name": "DeepSeek-V3-0324 card", "title": "deepseek-ai/DeepSeek-V3-0324 model card", "lab": "deepseek", "date": "2025-03-25", "printed": "49.2", "where": "card sentence \"LiveCodeBench: 39.2 → 49.2 (+10.0)\"" } ], "source_image": "https://frontierlog.ai/sources/0516abeb7983c595.webp", "source_image_original": "https://cdn.deepseek.com/api-docs/v3_0324_benchmark.webp" }, { "id": "gemini-2-5-pro:livecodebench-no-release:custom-window:1", "model": "gemini-2-5-pro", "model_name": "Gemini 2.5 Pro", "model_date": "2025-03-25", "event": "https://frontierlog.ai/event/2025-03-25-gemini-2-5-pro", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 69.0, "unit": "%", "printed": "69.0%", "reported_by": "google", "run_by": null, "label": "Gemini 2.5 Pro Thinking", "source": "https://blog.google/products-and-platforms/products/gemini/gemini-2-5-model-family-expands/", "source_title": "We’re expanding our Gemini 2.5 family of models", "source_date": "2025-06-17", "where": "benchmark table (image), row 'Code generation, LiveCodeBench (UI: 1/1/2025-5/1/2025)', column 'Gemini 2.5 Pro Thinking'", "own": false, "primary": true, "parallel": false, "notes": "thinking; pass@1; problems from 1/1/2025 to 5/1/2025", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://blog.google/products-and-platforms/products/gemini/gemini-2-5-pro-latest-preview/", "name": "Google post", "title": "Try the latest Gemini 2.5 Pro before general availability.", "lab": "google", "date": "2025-06-05", "printed": "69.0%", "where": "Benchmark table image, row Code generation LiveCodeBench (UI: 1/1/2025-5/1/2025) single attempt, column Gemini 2.5 Pro Preview 06-05 Thinking (the 06-05 preview, not labelled GA)" } ], "source_image": "https://frontierlog.ai/sources/4002c58e42db1f64.gif", "source_image_original": "https://storage.googleapis.com/gweb-uniblog-publish-prod/original_images/gemini_2-5_benchmarks_margin_light2x_1.gif" }, { "id": "qwen2-5-omni:livecodebench-no-release:custom-window:1", "model": "qwen2-5-omni", "model_name": "Qwen2.5-Omni", "model_date": "2025-03-27", "event": "https://frontierlog.ai/event/2025-03-27-qwen2-5-omni", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 24.6, "unit": "%", "printed": "24.6", "reported_by": "alibaba", "run_by": null, "label": "Qwen2.5-Omni-7B", "source": "https://huggingface.co/Qwen/Qwen2.5-Omni-7B/blob/5e19b3eabbf54ea08b3fcc2a2fcc6db3c6a7e09a/README.md", "source_title": "Qwen/Qwen2.5-Omni-7B model card (launch-day revision)", "source_date": "2025-03-27", "where": "model card 'Performance' section, 'Text -> Text' table, row 'LiveCodeBench2305-2409', column 'Qwen2.5-Omni-7B'", "own": true, "primary": true, "parallel": false, "notes": "problems from May 2023 to Sep 2024", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-28", "check": "human", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://github.com/QwenLM/Qwen2.5-Omni/blob/main/README.md", "name": "Qwen2.5-Omni card", "title": "Qwen2.5-Omni README", "lab": "alibaba", "date": "2025-03-27", "printed": "24.6", "where": "Text -> Text table, row LiveCodeBench 2305-2409, column Qwen2.5-Omni-7B" }, { "url": "https://github.com/QwenLM/Qwen2.5-Omni/blob/main/assets/Qwen2.5_Omni.pdf", "name": "Qwen2.5-Omni card", "title": "Qwen2.5-Omni Technical Report", "lab": "alibaba", "date": "2025-03-27", "printed": "24.6", "where": "Qwen2.5-Omni Technical Report, Table 1, row LiveCodeBench2305-2409, column Qwen2.5-Omni-7B" } ], "source_capture": "https://frontierlog.ai/captures/8124cb4cb84fec5c.png" }, { "id": "llama-4-maverick:livecodebench-no-release:custom-window:1", "model": "llama-4-maverick", "model_name": "Llama 4 Maverick", "model_date": "2025-04-05", "event": "https://frontierlog.ai/event/2025-04-05-llama-4", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 43.4, "unit": "%", "printed": "43.4", "reported_by": "meta", "run_by": null, "label": "Llama 4 Maverick", "source": "https://ai.meta.com/blog/llama-4-multimodal-intelligence/", "source_title": "The Llama 4 herd: The beginning of a new era of natively multimodal AI innovation", "source_date": "2025-04-05", "where": "'Llama 4 Maverick instruction-tuned benchmarks' table (image), row 'LiveCodeBench (10/01/2024-02/01/2025)', column 'Llama 4 Maverick'", "own": true, "primary": true, "parallel": false, "notes": "problems from Oct 2024 to Feb 2025; 0-shot; averaged over multiple generations", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-20", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/d5c84c16784d5728.png", "source_image_original": "https://scontent-ord5-2.xx.fbcdn.net/v/t39.2365-6/488688605_1406312723692874_1536535503366996614_n.png?_nc_cat=105&ccb=1-7&_nc_sid=e280be&_nc_ohc=ivZO1ZWL-4QQ7kNvwG2wTyx&_nc_oc=AdqN-GapDM5z68UF1nqjWnGufhFO_xDNGZIY6ht6MsXi0Jkyw7rJFBIRkzUkXXZXb7M&_nc_zt=14&_nc_ht=scontent-ord5-2.xx&_nc_gid=pzE_LW4H_OG--eyEc4cgFA&_nc_ss=7f289&oh=00_AQPJ4I6NWnMSCSGKckvJfnDqXojRhrLeJXSgjH0PfPVNCA&oe=6AF0150F" }, { "id": "llama-4-scout:livecodebench-no-release:custom-window:1", "model": "llama-4-scout", "model_name": "Llama 4 Scout", "model_date": "2025-04-05", "event": "https://frontierlog.ai/event/2025-04-05-llama-4", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 32.8, "unit": "%", "printed": "32.8", "reported_by": "meta", "run_by": null, "label": "Llama 4 Scout", "source": "https://ai.meta.com/blog/llama-4-multimodal-intelligence/", "source_title": "The Llama 4 herd: The beginning of a new era of natively multimodal AI innovation", "source_date": "2025-04-05", "where": "'Llama 4 Scout instruction-tuned benchmarks' table (image), row 'LiveCodeBench (10/01/2024-02/01/2025)', column 'Llama 4 Scout'", "own": true, "primary": true, "parallel": false, "notes": "problems from Oct 2024 to Feb 2025; 0-shot; averaged over multiple generations", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "double_entry", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/47ea0cbe5c73836d.png", "source_image_original": "https://scontent-ord5-1.xx.fbcdn.net/v/t39.2365-6/488658055_1347378876402143_3412007366291908454_n.png?_nc_cat=109&ccb=1-7&_nc_sid=e280be&_nc_ohc=OBma4RowFAYQ7kNvwEDKyuH&_nc_oc=AdqPMOh8lXBK3lg7quc_HNoP-49jOvB8JczkL0IsoavEpo4Igwy0r1vrenrHwo1uFmA&_nc_zt=14&_nc_ht=scontent-ord5-1.xx&_nc_gid=pzE_LW4H_OG--eyEc4cgFA&_nc_ss=7f289&oh=00_AQNipOGf-EGDwZJ5T-zywEffyFQYUnzut_O2CKWGuWBPNg&oe=6AF0271F" }, { "id": "gemini-2-5-flash:livecodebench-no-release:custom-window:1", "model": "gemini-2-5-flash", "model_name": "Gemini 2.5 Flash", "model_date": "2025-04-17", "event": "https://frontierlog.ai/event/2025-04-17-gemini-2-5-flash", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 41.1, "unit": "%", "printed": "41.1%", "reported_by": "google", "run_by": null, "label": "Gemini 2.5 Flash Non-thinking", "source": "https://blog.google/products-and-platforms/products/gemini/gemini-2-5-model-family-expands/", "source_title": "We’re expanding our Gemini 2.5 family of models", "source_date": "2025-06-17", "where": "benchmark table (image), row 'Code generation, LiveCodeBench (UI: 1/1/2025-5/1/2025)', column 'Gemini 2.5 Flash Non-thinking'", "own": false, "primary": false, "parallel": false, "notes": "non-thinking; pass@1; problems from 1/1/2025 to 5/1/2025", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-29", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/4002c58e42db1f64.gif", "source_image_original": "https://storage.googleapis.com/gweb-uniblog-publish-prod/original_images/gemini_2-5_benchmarks_margin_light2x_1.gif" }, { "id": "gemini-2-5-flash:livecodebench-no-release:custom-window:2", "model": "gemini-2-5-flash", "model_name": "Gemini 2.5 Flash", "model_date": "2025-04-17", "event": "https://frontierlog.ai/event/2025-04-17-gemini-2-5-flash", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 55.4, "unit": "%", "printed": "55.4%", "reported_by": "google", "run_by": null, "label": "Gemini 2.5 Flash Thinking", "source": "https://blog.google/products-and-platforms/products/gemini/gemini-2-5-model-family-expands/", "source_title": "We’re expanding our Gemini 2.5 family of models", "source_date": "2025-06-17", "where": "benchmark table (image), row 'Code generation, LiveCodeBench (UI: 1/1/2025-5/1/2025)', column 'Gemini 2.5 Flash Thinking'", "own": false, "primary": true, "parallel": false, "notes": "thinking; pass@1; problems from 1/1/2025 to 5/1/2025", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-29", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/4002c58e42db1f64.gif", "source_image_original": "https://storage.googleapis.com/gweb-uniblog-publish-prod/original_images/gemini_2-5_benchmarks_margin_light2x_1.gif" }, { "id": "gemini-2-5-flash:livecodebench-no-release:custom-window:3", "model": "gemini-2-5-flash", "model_name": "Gemini 2.5 Flash", "model_date": "2025-04-17", "event": "https://frontierlog.ai/event/2025-04-17-gemini-2-5-flash", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 59.3, "unit": "%", "printed": "59.3%", "reported_by": "google", "run_by": null, "label": "Gemini 2.5 Flash", "source": "https://storage.googleapis.com/deepmind-media/gemini/gemini_v2_5_report.pdf", "source_title": "Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities", "source_date": "2025-06-17", "where": "Table 3 (core capability benchmarks), row 'Code, LiveCodeBench', column 'Gemini 2.5 Flash'", "own": false, "primary": false, "parallel": false, "notes": "pass@1; single attempt; problems from 1/1/2025 to 5/1/2025", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-29", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [] }, { "id": "deepseek-r1-0528:livecodebench-no-release:custom-window:1", "model": "deepseek-r1-0528", "model_name": "DeepSeek-R1-0528", "model_date": "2025-05-28", "event": "https://frontierlog.ai/event/2025-05-28-deepseek-r1-0528", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 73.3, "unit": "%", "printed": "73.3", "reported_by": "deepseek", "run_by": null, "label": "DeepSeek-R1-0528", "source": "https://api-docs.deepseek.com/news/news250528/", "source_title": "DeepSeek-R1-0528 Release", "source_date": "2025-05-28", "where": "benchmark bar chart (image), group 'LiveCodeBench (Pass@1)', value label on the 'DeepSeek-R1-0528' bar", "own": true, "primary": true, "parallel": false, "notes": "problems from Aug 2024 to May 2025; averaged over 16 samples", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://huggingface.co/deepseek-ai/DeepSeek-R1-0528", "name": "DeepSeek-R1-0528 card", "title": "deepseek-ai/DeepSeek-R1-0528 model card", "lab": "deepseek", "date": "2025-05-28", "printed": "73.3", "where": "Evaluation table, row \"LiveCodeBench (2408-2505) (Pass@1)\", column \"DeepSeek R1 0528\"" }, { "url": "https://huggingface.co/deepseek-ai/DeepSeek-V3.1", "name": "DeepSeek-V3.1 card", "title": "deepseek-ai/DeepSeek-V3.1 model card", "lab": "deepseek", "date": "2025-08-21", "printed": "73.3", "where": "Evaluation table, column \"DeepSeek R1 0528\", row \"LiveCodeBench (2408-2505) (Pass@1)\"" } ], "source_image": "https://frontierlog.ai/sources/383ee2195827e346.png", "source_image_original": "https://cdn.deepseek.com/api-docs/r1_0528_benchmark.PNG" }, { "id": "minimax-m1:livecodebench-no-release:custom-window:1", "model": "minimax-m1", "model_name": "MiniMax-M1", "model_date": "2025-06-16", "event": "https://frontierlog.ai/event/2025-06-16-minimax-m1", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 65.0, "unit": "%", "printed": "65.0", "reported_by": "minimax", "run_by": null, "label": "MiniMax-M1-80k", "source": "https://www.minimax.io/news/minimaxm1", "source_title": "MiniMax-M1, the World's First Open-Source, Large-Scale, Hybrid-Attention Reasoning Model", "source_date": "2025-06-16", "where": "full results table image (last image, after the bullet list), row 'LiveCodeBench (24/8~25/5)', column 'MiniMax-M1-80k'", "own": true, "primary": true, "parallel": false, "notes": "MiniMax-M1-80k, 80K thinking budget; problems from Aug 2024 to May 2025", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-19", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/52957ae223f9c494.png", "source_image_original": "https://filecdn.minimax.chat/public/d50018df-d646-40e3-824b-e507c209ad03.png" }, { "id": "minimax-m1:livecodebench-no-release:custom-window:2", "model": "minimax-m1", "model_name": "MiniMax-M1", "model_date": "2025-06-16", "event": "https://frontierlog.ai/event/2025-06-16-minimax-m1", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 62.3, "unit": "%", "printed": "62.3", "reported_by": "minimax", "run_by": null, "label": "MiniMax-M1-40k", "source": "https://www.minimax.io/news/minimaxm1", "source_title": "MiniMax-M1, the World's First Open-Source, Large-Scale, Hybrid-Attention Reasoning Model", "source_date": "2025-06-16", "where": "full results table image (last image, after the bullet list), row 'LiveCodeBench (24/8~25/5)', column 'MiniMax-M1-40k'", "own": true, "primary": false, "parallel": false, "notes": "MiniMax-M1-40k, 40K thinking budget; problems from Aug 2024 to May 2025", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "double_entry", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/52957ae223f9c494.png", "source_image_original": "https://filecdn.minimax.chat/public/d50018df-d646-40e3-824b-e507c209ad03.png" }, { "id": "gemini-2-5-flash-lite:livecodebench-no-release:custom-window:1", "model": "gemini-2-5-flash-lite", "model_name": "Gemini 2.5 Flash-Lite", "model_date": "2025-06-17", "event": "https://frontierlog.ai/event/2025-06-17-gemini-2-5-ga-and-2-5-flash", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 33.7, "unit": "%", "printed": "33.7%", "reported_by": "google", "run_by": null, "label": "Gemini 2.5 Flash-Lite Preview 06-17 Non-thinking", "source": "https://blog.google/products-and-platforms/products/gemini/gemini-2-5-model-family-expands/", "source_title": "We’re expanding our Gemini 2.5 family of models", "source_date": "2025-06-17", "where": "benchmark table (image), row 'Code generation, LiveCodeBench', column 'Gemini 2.5 Flash-Lite Preview 06-17 Non-thinking'", "own": true, "primary": false, "parallel": false, "notes": "non-thinking; pass@1; problems from 1/1/2025 to 5/1/2025", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-26", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/4002c58e42db1f64.gif", "source_image_original": "https://storage.googleapis.com/gweb-uniblog-publish-prod/original_images/gemini_2-5_benchmarks_margin_light2x_1.gif" }, { "id": "gemini-2-5-flash-lite:livecodebench-no-release:custom-window:2", "model": "gemini-2-5-flash-lite", "model_name": "Gemini 2.5 Flash-Lite", "model_date": "2025-06-17", "event": "https://frontierlog.ai/event/2025-06-17-gemini-2-5-ga-and-2-5-flash", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 34.3, "unit": "%", "printed": "34.3%", "reported_by": "google", "run_by": null, "label": "Gemini 2.5 Flash-Lite Preview 06-17 Thinking", "source": "https://blog.google/products-and-platforms/products/gemini/gemini-2-5-model-family-expands/", "source_title": "We’re expanding our Gemini 2.5 family of models", "source_date": "2025-06-17", "where": "benchmark table (image), row 'Code generation, LiveCodeBench', column 'Gemini 2.5 Flash-Lite Preview 06-17 Thinking'", "own": true, "primary": true, "parallel": false, "notes": "thinking; pass@1; problems from 1/1/2025 to 5/1/2025", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-26", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/4002c58e42db1f64.gif", "source_image_original": "https://storage.googleapis.com/gweb-uniblog-publish-prod/original_images/gemini_2-5_benchmarks_margin_light2x_1.gif" }, { "id": "glm-4-5:livecodebench-no-release:custom-window:1", "model": "glm-4-5", "model_name": "GLM-4.5", "model_date": "2025-07-28", "event": "https://frontierlog.ai/event/2025-07-28-glm-4-5", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 72.9, "unit": "%", "printed": "72.9", "reported_by": "zai", "run_by": null, "label": "GLM-4.5", "source": "https://z.ai/blog/glm-4.5", "source_title": "GLM-4.5: Reasoning, Coding, and Agentic Abililties", "source_date": "2025-07-28", "where": "'Reasoning' section table, row 'LiveCodeBench 2407-2501', column 'GLM-4.5'", "own": true, "primary": true, "parallel": false, "notes": "thinking mode; problems from Jul 2024 to Jan 2025", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-19", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_capture": "https://frontierlog.ai/captures/9690be94d5b435fb.png" }, { "id": "glm-4-5-air:livecodebench-no-release:custom-window:1", "model": "glm-4-5-air", "model_name": "GLM-4.5-Air", "model_date": "2025-07-28", "event": "https://frontierlog.ai/event/2025-07-28-glm-4-5", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 70.7, "unit": "%", "printed": "70.7", "reported_by": "zai", "run_by": null, "label": "GLM-4.5-Air", "source": "https://z.ai/blog/glm-4.5", "source_title": "GLM-4.5: Reasoning, Coding, and Agentic Abililties", "source_date": "2025-07-28", "where": "Reasoning table (HTML), row 'LiveCodeBench 2407-2501', column 'GLM-4.5-Air'", "own": true, "primary": true, "parallel": false, "notes": "thinking mode; problems from Jul 2024 to Jan 2025", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "double_entry", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_capture": "https://frontierlog.ai/captures/686f303d4261e1c1.png" }, { "id": "llama-nemotron-super-49b-v1-5:livecodebench-no-release:custom-window:1", "model": "llama-nemotron-super-49b-v1-5", "model_name": "Llama Nemotron Super 49B v1.5", "model_date": "2025-07-29", "event": "https://frontierlog.ai/event/2025-07-29-llama-nemotron-super-v1-5", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 73.6, "unit": "%", "printed": "73.6", "reported_by": "nvidia", "run_by": null, "label": "Llama Nemotron Super 49B v1.5", "source": "https://developer.nvidia.com/blog/build-more-accurate-and-efficient-ai-agents-with-the-new-nvidia-llama-nemotron-super-v1-5/", "source_title": "Build More Accurate and Efficient AI Agents with the New NVIDIA Llama Nemotron Super v1.5", "source_date": "2025-07-29", "where": "Figure 2 bar chart (image), value label of the 'Llama Nemotron Super 49B v1.5' bar (green, leftmost in group) in group 'LiveCodeBench'", "own": true, "primary": true, "parallel": false, "notes": "problems from Oct 2024 to Feb 2025", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://huggingface.co/nvidia/Llama-3_3-Nemotron-Super-49B-v1_5/blob/420ba7d28211abf116b8b103ab700d92619daf98/README.md", "name": "Llama Nemotron Super v1.5 card", "title": "nvidia/Llama-3_3-Nemotron-Super-49B-v1_5 model card (revision of 2025-10-15)", "lab": "nvidia", "date": "2025-07-29", "printed": "73.58", "where": "Evaluation Results, 'LiveCodeBench 24.10-25.02' table, Reasoning On, pass@1 (avg. over 4 runs)" } ], "source_image": "https://frontierlog.ai/sources/d431180dea13f63f.png", "source_image_original": "https://developer-blogs.nvidia.com/wp-content/uploads/2025/07/image-29.png" }, { "id": "nemotron-nano-9b-v2:livecodebench-no-release:window-unstated:1", "model": "nemotron-nano-9b-v2", "model_name": "Nemotron Nano 9B v2", "model_date": "2025-08-18", "event": "https://frontierlog.ai/event/2025-08-18-nemotron-nano-2", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 71.1, "unit": "%", "printed": "71.1", "reported_by": "nvidia", "run_by": null, "label": "NVIDIA-Nemotron-Nano-9B-v2", "source": "https://research.nvidia.com/labs/adlr/NVIDIA-Nemotron-Nano-2/", "source_title": "NVIDIA Nemotron Nano 2 and the Nemotron Pretraining Dataset v1", "source_date": "2025-08-18", "where": "model comparison bar chart (image, 'Measured Accuracy'), value label of the 'NVIDIA-Nemotron-Nano-9B-v2' bar (green) in group 'LiveCodeBench (Coding)'", "own": true, "primary": true, "parallel": false, "notes": "reasoning on", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://huggingface.co/nvidia/NVIDIA-Nemotron-Nano-9B-v2-FP8/blob/8bc5eece2eb5514c4bca7f2ec655b91eb554f4c0/README.md", "name": "Nemotron Nano 2 card", "title": "nvidia/NVIDIA-Nemotron-Nano-9B-v2-FP8 model card (revision of 2026-01-08)", "lab": "nvidia", "date": "2025-08-18", "printed": "71.1%", "where": "FP8 model card, Benchmark Results (Reasoning On) table, LCB row, NVIDIA-Nemotron-Nano-9B-v2 (BF16) column" }, { "url": "https://huggingface.co/nvidia/NVIDIA-Nemotron-Nano-9B-v2-NVFP4/blob/8556c9164ddb43fe1f4f4ad730593b3c5e3f7328/README.md", "name": "Nemotron Nano 2 card", "title": "nvidia/NVIDIA-Nemotron-Nano-9B-v2-NVFP4 model card (revision of 2026-01-08)", "lab": "nvidia", "date": "2025-08-18", "printed": "71.1%", "where": "NVFP4 model card, Benchmark Results (Reasoning On) table, LCB row, NVIDIA-Nemotron-Nano-9B-v2 (BF16) column" }, { "url": "https://huggingface.co/nvidia/NVIDIA-Nemotron-Nano-9B-v2/blob/6533e8de2c68e4536bf7c411d7a3ce5734111476/README.md", "name": "Nemotron Nano 2 card", "title": "nvidia/NVIDIA-Nemotron-Nano-9B-v2 model card (revision of 2026-03-05)", "lab": "nvidia", "date": "2025-08-18", "printed": "71.1%", "where": "Benchmark Results (Reasoning On) table, LCB row, NVIDIA-Nemotron-Nano-9B-v2 column" }, { "url": "https://research.nvidia.com/labs/adlr/files/NVIDIA-Nemotron-Nano-2-Technical-Report.pdf", "name": "Nemotron Nano 2 card", "title": "NVIDIA Nemotron Nano 2 technical report", "lab": "nvidia", "date": "2025-08-18", "printed": "71.1", "where": "Figure 1 (Measured Accuracy), LiveCodeBench bar, NVIDIA-Nemotron-Nano-9B-v2 (vector text in PDF)" } ], "source_image": "https://frontierlog.ai/sources/5d46904619c03524.png", "source_image_original": "https://research.nvidia.com/labs/adlr/images/nvidia-nemotron-nano-2/nano-2-model-comparison.png" }, { "id": "deepseek-v3-1:livecodebench-no-release:window-unstated:1", "model": "deepseek-v3-1", "model_name": "DeepSeek-V3.1", "model_date": "2025-08-21", "event": "https://frontierlog.ai/event/2025-08-21-deepseek-v3-1", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 74.8, "unit": "%", "printed": "74.8%", "reported_by": "deepseek", "run_by": null, "label": "V3.1-Think", "source": "https://api-docs.deepseek.com/news/news250821/", "source_title": "DeepSeek-V3.1 Release", "source_date": "2025-08-21", "where": "third image (thinking efficiency bar chart), group 'LiveCodeBench', hatched bar 'V3.1-Think', label above the bar (output tokens, then the score in parentheses)", "own": true, "primary": true, "parallel": false, "notes": "thinking mode; 13,977 output tokens on average", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-19", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/46ed66f55ee72932.webp", "source_image_original": "https://cdn.deepseek.com/api-docs/v3.1_benchmark_3.webp" }, { "id": "deepseek-v3-1:livecodebench-no-release:custom-window:1", "model": "deepseek-v3-1", "model_name": "DeepSeek-V3.1", "model_date": "2025-08-21", "event": "https://frontierlog.ai/event/2025-08-21-deepseek-v3-1", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 56.4, "unit": "%", "printed": "56.4", "reported_by": "deepseek", "run_by": null, "label": "DeepSeek V3.1-NonThinking", "source": "https://huggingface.co/deepseek-ai/DeepSeek-V3.1", "source_title": "deepseek-ai/DeepSeek-V3.1 model card", "source_date": "2025-08-21", "where": "model card, Evaluation table, row 'LiveCodeBench (2408-2505) (Pass@1)', column 'DeepSeek V3.1-NonThinking'", "own": true, "primary": true, "parallel": false, "notes": "non-thinking mode; problems from Aug 2024 to May 2025", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-29", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_capture": "https://frontierlog.ai/captures/40188a910b83f710.png" }, { "id": "longcat-flash-chat:livecodebench-no-release:window-unstated:1", "model": "longcat-flash-chat", "model_name": "LongCat-Flash-Chat", "model_date": "2025-09-01", "event": "https://frontierlog.ai/event/2025-09-01-longcat-flash-chat", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 48.02, "unit": "%", "printed": "48.02", "reported_by": "meituan", "run_by": null, "label": "LongCat-Flash", "source": "https://huggingface.co/meituan-longcat/LongCat-Flash-Chat", "source_title": "LongCat-Flash-Chat model card (Hugging Face)", "source_date": "2025-09-01", "where": "model card, Evaluation Results table, row 'LiveCodeBench (pass@1)', column 'LongCat-Flash'", "own": true, "primary": true, "parallel": false, "notes": "non-thinking", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://huggingface.co/meituan-longcat/LongCat-Flash-Omni", "name": "LongCat-Flash-Omni card", "title": "LongCat-Flash-Omni model card (Hugging Face)", "lab": "meituan", "date": "2025-11-03", "printed": "48.02", "where": "LongCat-Flash-Omni HF card, Text table, LiveCodeBench (pass@1) row, 'LongCat-Flash' column" }, { "url": "https://github.com/meituan-longcat/LongCat-Flash-Chat", "name": "LongCat-Flash-Chat card", "title": "LongCat-Flash-Chat GitHub repository", "lab": "meituan", "date": "2025-09-01", "printed": "48.02", "where": "GitHub README, evaluation table, LiveCodeBench (pass@1) row, LongCat-Flash column" }, { "url": "https://github.com/meituan-longcat/LongCat-Flash-Chat/blob/main/tech_report.pdf", "name": "LongCat-Flash-Chat card", "title": "LongCat-Flash Technical Report", "lab": "meituan", "date": "2025-09-01", "printed": "48.02", "where": "LongCat-Flash Technical Report (PDF in the lab's GitHub repo), chat-model evaluation table, LiveCodeBench (pass@1) row, LongCat-Flash column" }, { "url": "https://github.com/meituan-longcat/LongCat-Flash-Omni", "name": "LongCat-Flash-Omni card", "title": "LongCat-Flash-Omni GitHub repository", "lab": "meituan", "date": "2025-11-03", "printed": "48.02", "where": "LongCat-Flash-Omni GitHub README, Text table, LiveCodeBench (pass@1) row, 'LongCat-Flash' column" }, { "url": "https://github.com/meituan-longcat/LongCat-Flash-Omni/blob/main/tech_report.pdf", "name": "LongCat-Flash-Omni card", "title": "LongCat-Flash-Omni Technical Report", "lab": "meituan", "date": "2025-11-03", "printed": "48.02", "where": "LongCat-Flash-Omni Technical Report, Table 13 (frontier chat/instruct models), LiveCodeBench (pass@1) row, 'LongCat-Flash' column" } ], "source_capture": "https://frontierlog.ai/captures/b4e44a3a773f84f7.png" }, { "id": "deepseek-v3-1-terminus:livecodebench-no-release:custom-window:1", "model": "deepseek-v3-1-terminus", "model_name": "DeepSeek-V3.1-Terminus", "model_date": "2025-09-22", "event": "https://frontierlog.ai/event/2025-09-22-deepseek-v3-1-terminus", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 74.9, "unit": "%", "printed": "74.9", "reported_by": "deepseek", "run_by": null, "label": "DeepSeek-V3.1-Terminus", "source": "https://api-docs.deepseek.com/news/news250922/", "source_title": "DeepSeek-V3.1-Terminus", "source_date": "2025-09-22", "where": "benchmark table (image), row 'LiveCodeBench', column 'DeepSeek-V3.1-Terminus'", "own": true, "primary": true, "parallel": false, "notes": "reasoning mode without tools; problems from Aug 2024 to May 2025", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://huggingface.co/deepseek-ai/DeepSeek-V3.1-Terminus", "name": "DeepSeek-V3.1-Terminus card", "title": "deepseek-ai/DeepSeek-V3.1-Terminus model card", "lab": "deepseek", "date": "2025-09-22", "printed": "74.9", "where": "Benchmark table, column \"DeepSeek-V3.1-Terminus\", row \"LiveCodeBench\"" }, { "url": "https://huggingface.co/deepseek-ai/DeepSeek-V3.2-Exp", "name": "DeepSeek-V3.2-Exp card", "title": "deepseek-ai/DeepSeek-V3.2-Exp model card", "lab": "deepseek", "date": "2025-09-29", "printed": "74.9", "where": "Benchmark table, column \"DeepSeek-V3.1-Terminus\", row \"LiveCodeBench\"" }, { "url": "https://api-docs.deepseek.com/news/news250929/", "name": "DeepSeek-V3.2-Exp", "title": "Introducing DeepSeek-V3.2-Exp", "lab": "deepseek", "date": "2025-09-29", "printed": "74.9", "where": "benchmark image, column \"DeepSeek-V3.1-Terminus\", row \"LiveCodeBench\"" }, { "url": "https://github.com/deepseek-ai/DeepSeek-V3.2-Exp/blob/main/DeepSeek_V3_2.pdf", "name": "DeepSeek-V3.2-Exp card", "title": "DeepSeek-V3.2-Exp: Boosting Long-Context Efficiency with DeepSeek Sparse Attention", "lab": "deepseek", "date": "2025-09-29", "printed": "74.9", "where": "Table 1, column \"DeepSeek-V3.1-Terminus\", row \"LiveCodeBench (2408-2505) (Pass@1)\"" } ], "source_image": "https://frontierlog.ai/sources/93e083b75de124e9.png", "source_image_original": "https://api-docs.deepseek.com/zh-cn/img/deepseek_v3.1_terminus_benchmark_en.png" }, { "id": "longcat-flash-thinking:livecodebench-no-release:custom-window:1", "model": "longcat-flash-thinking", "model_name": "LongCat-Flash-Thinking", "model_date": "2025-09-22", "event": "https://frontierlog.ai/event/2025-09-22-longcat-flash-thinking", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 79.4, "unit": "%", "printed": "79.4", "reported_by": "meituan", "run_by": null, "label": "LongCat-Flash-Thinking", "source": "https://tech.meituan.com/2025/09/22/LongCat-Flash-Thinking.html", "source_title": "LongCat-Flash-Thinking 正式发布，更强、更专业，保持极速！", "source_date": "2025-09-22", "where": "capability list, coding item (sentence on LiveCodeBench); the chart labels the window 24.08-25.05", "own": true, "primary": true, "parallel": false, "notes": "average of 4 runs; problems from Aug 2024 to May 2025", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://huggingface.co/meituan-longcat/LongCat-Flash-Thinking", "name": "LongCat-Flash-Thinking card", "title": "LongCat-Flash-Thinking model card (Hugging Face)", "lab": "meituan", "date": "2025-09-22", "printed": "79.4", "where": "Evaluation table, LiveCodeBench (Mean@4) row, LongCat-Flash-Thinking column" }, { "url": "https://github.com/meituan-longcat/LongCat-Flash-Thinking", "name": "LongCat-Flash-Thinking card", "title": "LongCat-Flash-Thinking GitHub repository", "lab": "meituan", "date": "2025-09-22", "printed": "79.4", "where": "GitHub README, evaluation table, LiveCodeBench (Mean@4) row, LongCat-Flash-Thinking column" }, { "url": "https://github.com/meituan-longcat/LongCat-Flash-Thinking/blob/main/tech_report.pdf", "name": "LongCat-Flash-Thinking card", "title": "LongCat-Flash-Thinking Technical Report", "lab": "meituan", "date": "2025-09-22", "printed": "79.4", "where": "LongCat-Flash-Thinking Technical Report (PDF in GitHub repo), main evaluation table, LCB (24.08-25.05)(Mean@4) row, LongCat-Flash-Thinking column" } ] }, { "id": "deepseek-v3-2-exp:livecodebench-no-release:custom-window:1", "model": "deepseek-v3-2-exp", "model_name": "DeepSeek-V3.2-Exp", "model_date": "2025-09-29", "event": "https://frontierlog.ai/event/2025-09-29-deepseek-v3-2-exp", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 74.1, "unit": "%", "printed": "74.1", "reported_by": "deepseek", "run_by": null, "label": "DeepSeek-V3.2-Exp", "source": "https://api-docs.deepseek.com/news/news250929/", "source_title": "Introducing DeepSeek-V3.2-Exp", "source_date": "2025-09-29", "where": "benchmark table (image), Code section, row 'LiveCodeBench', column 'DeepSeek-V3.2-Exp'", "own": true, "primary": true, "parallel": false, "notes": "reasoning mode without tools; problems from Aug 2024 to May 2025", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://huggingface.co/deepseek-ai/DeepSeek-V3.2-Exp", "name": "DeepSeek-V3.2-Exp card", "title": "deepseek-ai/DeepSeek-V3.2-Exp model card", "lab": "deepseek", "date": "2025-09-29", "printed": "74.1", "where": "Benchmark table, column \"DeepSeek-V3.2-Exp\", row \"LiveCodeBench\"" }, { "url": "https://github.com/deepseek-ai/DeepSeek-V3.2-Exp/blob/main/DeepSeek_V3_2.pdf", "name": "DeepSeek-V3.2-Exp card", "title": "DeepSeek-V3.2-Exp: Boosting Long-Context Efficiency with DeepSeek Sparse Attention", "lab": "deepseek", "date": "2025-09-29", "printed": "74.1", "where": "Table 1, column \"DeepSeek-V3.2-Exp\", row \"LiveCodeBench (2408-2505) (Pass@1)\"" } ], "source_image": "https://frontierlog.ai/sources/4d9152a1f18ce8ee.webp", "source_image_original": "https://api-docs.deepseek.com/img/v3_2_benchmark.webp" }, { "id": "minimax-m2:livecodebench-no-release:window-unstated:1", "model": "minimax-m2", "model_name": "MiniMax M2", "model_date": "2025-10-27", "event": "https://frontierlog.ai/event/2025-10-27-minimax-m2", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 83.0, "unit": "%", "printed": "83.0", "reported_by": "minimax", "run_by": null, "label": "MiniMax-M2", "source": "https://www.minimax.io/news/minimax-m21", "source_title": "MiniMax M2.1: Significantly Enhanced Multi-Language Programming, Built for Real-World Complex Tasks", "source_date": "2025-12-23", "where": "later report: MiniMax-M2.1 post, 'Benchmarks' section, last (long-horizon tool use and intelligence) table image, row 'LCB', column 'MiniMax-M2'", "own": false, "primary": true, "parallel": false, "notes": "internal run following Artificial Analysis Intelligence Index methods", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://huggingface.co/MiniMaxAI/MiniMax-M2.1", "name": "MiniMax M2.1 card", "title": "MiniMaxAI/MiniMax-M2.1 model card", "lab": "minimax", "date": "2025-12-23", "printed": "83.0", "where": "M2.1 card AA-method table, column \"MiniMax-M2\", row \"LCB\"" }, { "url": "https://github.com/MiniMax-AI/MiniMax-M2.1", "name": "MiniMax M2.1 card", "title": "MiniMax-M2.1 GitHub repository", "lab": "minimax", "date": "2025-12-23", "printed": "83.0", "where": "M2.1 README AA-method table, column \"MiniMax-M2\", row \"LCB\"" } ], "source_image": "https://frontierlog.ai/sources/78a04f05b9d1589b.png", "source_image_original": "https://file.cdn.minimax.io/public/6aaadb26-4395-46ba-9ebe-f858f3bc0c6b.png" }, { "id": "longcat-flash-omni:livecodebench-no-release:window-unstated:1", "model": "longcat-flash-omni", "model_name": "LongCat-Flash-Omni", "model_date": "2025-11-03", "event": "https://frontierlog.ai/event/2025-11-03-longcat-flash-omni", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 52.64, "unit": "%", "printed": "52.64", "reported_by": "meituan", "run_by": null, "label": "LongCat-Flash-Omni Instruct", "source": "https://huggingface.co/meituan-longcat/LongCat-Flash-Omni", "source_title": "LongCat-Flash-Omni model card (Hugging Face)", "source_date": "2025-11-03", "where": "model card, Text benchmark table (collapsed 'Text' section), row 'LiveCodeBench (pass@1)', column 'LongCat-Flash-Omni Instruct'", "own": true, "primary": true, "parallel": false, "notes": "", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://github.com/meituan-longcat/LongCat-Flash-Omni", "name": "LongCat-Flash-Omni card", "title": "LongCat-Flash-Omni GitHub repository", "lab": "meituan", "date": "2025-11-03", "printed": "52.64", "where": "GitHub README, Text table, LiveCodeBench (pass@1) row, LongCat-Flash-Omni Instruct column" }, { "url": "https://github.com/meituan-longcat/LongCat-Flash-Omni/blob/main/tech_report.pdf", "name": "LongCat-Flash-Omni card", "title": "LongCat-Flash-Omni Technical Report", "lab": "meituan", "date": "2025-11-03", "printed": "52.64", "where": "Technical Report (PDF in GitHub repo), Table 13, LiveCodeBench (pass@1) row, LongCat-Flash-Omni Instruct column" } ], "source_capture": "https://frontierlog.ai/captures/afb0fa5cde611f0c.png" }, { "id": "gigachat-3-lightning:livecodebench-no-release:window-unstated:1", "model": "gigachat-3-lightning", "model_name": "GigaChat 3 Lightning", "model_date": "2025-11-21", "event": "https://frontierlog.ai/event/2025-11-21-gigachat-3-ultra-preview", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 20.31, "unit": "%", "printed": "0,2031", "reported_by": "sber", "run_by": null, "label": "GigaChat-Lightning-sft", "source": "https://habr.com/ru/companies/sberdevices/articles/968904/", "source_title": "GigaChat 3 Ultra Preview — тяжёлый open source", "source_date": "2025-11-21", "where": "alignment metrics table (HTML, under 'Alignment-метрики'), row 'LiveCodeBench_ZERO_SHOT', column 'GigaChat-Lightning-sft'", "own": true, "primary": true, "parallel": false, "notes": "0-shot", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "double_entry", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_capture": "https://frontierlog.ai/captures/bde5428c2608d17f.png" }, { "id": "deepseek-v3-2:livecodebench-no-release:custom-window:1", "model": "deepseek-v3-2", "model_name": "DeepSeek-V3.2", "model_date": "2025-12-01", "event": "https://frontierlog.ai/event/2025-12-01-deepseek-v3-2", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 83.3, "unit": "%", "printed": "83.3", "reported_by": "deepseek", "run_by": null, "label": "DeepSeek-V3.2 Thinking", "source": "https://api-docs.deepseek.com/news/news251201/", "source_title": "DeepSeek-V3.2 Release", "source_date": "2025-12-01", "where": "reasoning table (image), row 'LiveCodeBench', column 'DeepSeek-V3.2 Thinking'", "own": true, "primary": true, "parallel": false, "notes": "thinking mode; 16K output tokens; problems from Aug 2024 to Apr 2025", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf", "name": "DeepSeek-V3.2 card", "title": "DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models", "lab": "deepseek", "date": "2025-12-01", "printed": "83.3", "where": "Table 2 and Table 3 (printed \"83.3 (16k)\"), column \"DeepSeek-V3.2 Thinking\", row \"LiveCodeBench (Pass@1-COT)\"; problems 2024.08-2025.04" } ], "source_image": "https://frontierlog.ai/sources/a313f1f47e0fe2c0.png", "source_image_original": "https://api-docs.deepseek.com/img/v3.2_251201_benchmark_table_en.png" }, { "id": "deepseek-v3-2-speciale:livecodebench-no-release:custom-window:1", "model": "deepseek-v3-2-speciale", "model_name": "DeepSeek-V3.2-Speciale", "model_date": "2025-12-01", "event": "https://frontierlog.ai/event/2025-12-01-deepseek-v3-2", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 88.7, "unit": "%", "printed": "88.7", "reported_by": "deepseek", "run_by": null, "label": "DeepSeek-V3.2 Speciale", "source": "https://api-docs.deepseek.com/news/news251201/", "source_title": "DeepSeek-V3.2 Release", "source_date": "2025-12-01", "where": "reasoning table (image), row 'LiveCodeBench', column 'DeepSeek-V3.2 Speciale'", "own": true, "primary": true, "parallel": false, "notes": "27K output tokens; problems from Aug 2024 to Apr 2025", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf", "name": "DeepSeek-V3.2 card", "title": "DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models", "lab": "deepseek", "date": "2025-12-01", "printed": "88.7", "where": "Table 3, column \"DeepSeek-V3.2 Speciale\", row \"LiveCodeBench (Pass@1-COT)\", printed \"88.7 (27k)\"" } ], "source_image": "https://frontierlog.ai/sources/a313f1f47e0fe2c0.png", "source_image_original": "https://api-docs.deepseek.com/img/v3.2_251201_benchmark_table_en.png" }, { "id": "mistral-large-3:livecodebench-no-release:window-unstated:1", "model": "mistral-large-3", "model_name": "Mistral Large 3 675B", "model_date": "2025-12-02", "event": "https://frontierlog.ai/event/2025-12-02-mistral-3", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 34.4, "unit": "%", "printed": "34.4", "reported_by": "mistral", "run_by": null, "label": "Mistral Large 3 (675B)", "source": "https://mistral.ai/news/mistral-3/", "source_title": "Introducing Mistral 3", "source_date": "2025-12-02", "where": "'Base Model Benchmark Comparison' bar chart image, group 'LiveCodeBench (no CoT)', bar 'Mistral Large 3 (675B)', value label", "own": true, "primary": true, "parallel": false, "notes": "base model, before instruction tuning; no chain of thought", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-20", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/b74e86a313e7eaf2.webp", "source_image_original": "https://mistral.ai/_astro/98aeee04-e1c3-43b7-b90e-c51da84d5e56_19WeLG.webp" }, { "id": "minimax-m2-1:livecodebench-no-release:window-unstated:1", "model": "minimax-m2-1", "model_name": "MiniMax M2.1", "model_date": "2025-12-23", "event": "https://frontierlog.ai/event/2025-12-23-minimax-m2-1", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 78.0, "unit": "%", "printed": "78.0", "reported_by": "minimax", "run_by": null, "label": "MiniMax-M2.1", "source": "https://www.minimax.io/news/minimax-m21", "source_title": "MiniMax M2.1: Significantly Enhanced Multi-Language Programming, Built for Real-World Complex Tasks", "source_date": "2025-12-23", "where": "'Benchmarks' section, last (long-horizon tool use and intelligence) table image, row 'LCB', column 'MiniMax-M2.1'", "own": true, "primary": true, "parallel": false, "notes": "internal run following Artificial Analysis Intelligence Index methods", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "double_entry", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/78a04f05b9d1589b.png", "source_image_original": "https://file.cdn.minimax.io/public/6aaadb26-4395-46ba-9ebe-f858f3bc0c6b.png" }, { "id": "minimax-m2-1:livecodebench-no-release:window-unstated:2", "model": "minimax-m2-1", "model_name": "MiniMax M2.1", "model_date": "2025-12-23", "event": "https://frontierlog.ai/event/2025-12-23-minimax-m2-1", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 81.0, "unit": "%", "printed": "81.0", "reported_by": "minimax", "run_by": null, "label": "MiniMax-M2.1", "source": "https://huggingface.co/MiniMaxAI/MiniMax-M2.1", "source_title": "MiniMaxAI/MiniMax-M2.1 model card", "source_date": "2025-12-23", "where": "model card, long-horizon tool use and intelligence table, row 'LCB', column 'MiniMax-M2.1'", "own": true, "primary": false, "parallel": false, "notes": "internal run following Artificial Analysis Intelligence Index methods", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://github.com/MiniMax-AI/MiniMax-M2.1", "name": "MiniMax M2.1 card", "title": "MiniMax-M2.1 GitHub repository", "lab": "minimax", "date": "2025-12-23", "printed": "81.0", "where": "README AA-method table, column \"MiniMax-M2.1\", row \"LCB\"" } ], "source_capture": "https://frontierlog.ai/captures/e2ae654583722dcb.png" }, { "id": "longcat-flash-thinking-2601:livecodebench-no-release:custom-window:1", "model": "longcat-flash-thinking-2601", "model_name": "LongCat-Flash-Thinking-2601", "model_date": "2026-01-20", "event": "https://frontierlog.ai/event/2026-01-20-longcat-flash-thinking-2601", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 82.8, "unit": "%", "printed": "82.8", "reported_by": "meituan", "run_by": null, "label": "LongCat-Flash-Thinking-2601", "source": "https://tech.meituan.com/2026/01/20/LongCat-Flash-Thinking-2601.html", "source_title": "美团 LongCat-Flash-Thinking-2601 发布，工具调用能力登顶开源 SOTA！", "source_date": "2026-01-20", "where": "capability list, coding item (sentence on LCB); the chart labels the window 24.08-25.05", "own": true, "primary": true, "parallel": false, "notes": "average of 4 runs; problems from Aug 2024 to May 2025", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://huggingface.co/meituan-longcat/LongCat-Flash-Thinking-2601", "name": "LongCat-Flash-Thinking-2601 card", "title": "LongCat-Flash-Thinking-2601 model card (Hugging Face)", "lab": "meituan", "date": "2026-01-20", "printed": "82.8", "where": "HF model card, LCB (24.08–25.05) (Avg@4) row, LongCat-Flash-Thinking-2601 column" }, { "url": "https://github.com/meituan-longcat/LongCat-Flash-Thinking-2601", "name": "LongCat-Flash-Thinking-2601 card", "title": "LongCat-Flash-Thinking-2601 GitHub repository", "lab": "meituan", "date": "2026-01-20", "printed": "82.8", "where": "GitHub README, LCB (24.08–25.05) (Avg@4) row, LongCat-Flash-Thinking-2601 column" }, { "url": "https://github.com/meituan-longcat/LongCat-Flash-Thinking-2601/blob/main/LongCat_Flash_Thinking_2601_Technical_Report.pdf", "name": "LongCat-Flash-Thinking-2601 card", "title": "LongCat-Flash-Thinking-2601 Technical Report", "lab": "meituan", "date": "2026-01-20", "printed": "82.8", "where": "Technical Report (PDF in GitHub repo), main evaluation table, LCB (24.08–25.05) (Avg@4) row, LongCat-Flash-Thinking-2601 column" } ] }, { "id": "gemini-3-1-flash-lite:livecodebench-no-release:custom-window:1", "model": "gemini-3-1-flash-lite", "model_name": "Gemini 3.1 Flash-Lite", "model_date": "2026-03-03", "event": "https://frontierlog.ai/event/2026-03-03-gemini-3-1-flash-lite", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "custom-window", "setting_name": "Lab-chosen window", "score": 72.0, "unit": "%", "printed": "72.0%", "reported_by": "google", "run_by": null, "label": "Gemini 3.1 Flash-Lite High", "source": "https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-1-flash-lite/", "source_title": "Gemini 3.1 Flash-Lite: Built for intelligence at scale", "source_date": "2026-03-03", "where": "benchmark table (image), row 'LiveCodeBench (UI: 1/1/2025-5/1/2025)', column 'Gemini 3.1 Flash-Lite High'", "own": true, "primary": true, "parallel": false, "notes": "high thinking; problems from 1/1/2025 to 5/1/2025", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-29", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/176bfed80b7bd3e1.gif", "source_image_original": "https://storage.googleapis.com/gweb-uniblog-publish-prod/original_images/gemini-3.1-flash-lite-table_1.gif" }, { "id": "mistral-small-4:livecodebench-no-release:window-unstated:1", "model": "mistral-small-4", "model_name": "Mistral Small 4 119B", "model_date": "2026-03-16", "event": "https://frontierlog.ai/event/2026-03-16-mistral-small-4", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 63.6, "unit": "%", "printed": "63.6", "reported_by": "mistral", "run_by": null, "label": "Mistral Small 4 - High", "source": "https://mistral.ai/news/mistral-small-4/", "source_title": "Introducing Mistral Small 4", "source_date": "2026-03-16", "where": "chart image 'Performance comparison across internal models' (second), group 'LiveCodeBench', bar 'Mistral Small 4 - High', value label", "own": true, "primary": true, "parallel": false, "notes": "high reasoning effort", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "double_entry", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/6a20cec3b0654df9.webp", "source_image_original": "https://mistral.ai/_astro/f6b89244-752a-45ce-8eed-166a58680399_138EpE.webp" }, { "id": "deepseek-v4-flash:livecodebench-no-release:window-unstated:1", "model": "deepseek-v4-flash", "model_name": "DeepSeek-V4-Flash", "model_date": "2026-04-24", "event": "https://frontierlog.ai/event/2026-04-24-deepseek-v4-preview", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 91.6, "unit": "%", "printed": "91.6", "reported_by": "deepseek", "run_by": null, "label": "DS-V4-Flash Max", "source": "https://api-docs.deepseek.com/news/news260424/", "source_title": "DeepSeek V4 Preview Release", "source_date": "2026-04-24", "where": "benchmark table (second image), row 'LiveCodeBench (Pass@1)', column 'DS-V4-Flash Max'", "own": true, "primary": true, "parallel": false, "notes": "max reasoning effort", "caveat": "", "human_verified": false, "human_verified_on": null, "check": "corroborated", "corroborated": true, "corroborated_on": "2026-10-28", "corroborated_by": [ { "url": "https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro", "name": "DeepSeek V4 Preview card", "title": "DeepSeek-V4-Pro model card", "lab": "deepseek", "date": "2026-04-24", "printed": "91.6", "where": "card instruct table (non-think/high/max), column \"V4-Flash Max\", row \"LiveCodeBench (Pass@1)\"" } ], "source_image": "https://frontierlog.ai/sources/9226203de8e24b33.png", "source_image_original": "https://api-docs.deepseek.com/img/v4-benchmark-2.png" }, { "id": "deepseek-v4-pro:livecodebench-no-release:window-unstated:1", "model": "deepseek-v4-pro", "model_name": "DeepSeek-V4-Pro", "model_date": "2026-04-24", "event": "https://frontierlog.ai/event/2026-04-24-deepseek-v4-preview", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 93.5, "unit": "%", "printed": "93.5", "reported_by": "deepseek", "run_by": null, "label": "DS-V4-Pro Max", "source": "https://api-docs.deepseek.com/news/news260424/", "source_title": "DeepSeek V4 Preview Release", "source_date": "2026-04-24", "where": "benchmark table image, row 'LiveCodeBench (Pass@1)', column 'DS-V4-Pro Max'", "own": true, "primary": true, "parallel": false, "notes": "max reasoning effort", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-18", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_image": "https://frontierlog.ai/sources/9226203de8e24b33.png", "source_image_original": "https://api-docs.deepseek.com/img/v4-benchmark-2.png" }, { "id": "qwen3-7-max:livecodebench-no-release:window-unstated:1", "model": "qwen3-7-max", "model_name": "Qwen3.7-Max", "model_date": "2026-05-20", "event": "https://frontierlog.ai/event/2026-05-20-qwen3-7-max", "benchmark": "livecodebench-no-release", "benchmark_name": "LiveCodeBench, no release named", "setting": "window-unstated", "setting_name": "Window not stated", "score": 91.6, "unit": "%", "printed": "91.6", "reported_by": "alibaba", "run_by": null, "label": "Qwen3.7-Max", "source": "https://qwen.ai/blog?id=qwen3.7", "source_title": "Qwen3.7: The Agent Frontier", "source_date": "2026-05-20", "where": "'Performance' table, row 'LiveCodeBench', column 'Qwen3.7-Max'", "own": true, "primary": true, "parallel": false, "notes": "", "caveat": "", "human_verified": true, "human_verified_on": "2026-10-19", "check": "human", "corroborated": false, "corroborated_on": null, "corroborated_by": [], "source_capture": "https://frontierlog.ai/captures/9d2174c2d8ef254a.png" } ] }