{
  "api_version": "1",
  "updated": "2026-10-27",
  "latest_event": "2026-10-26",
  "rights": "© Frontier Log, all rights reserved.",
  "total": 313,
  "events": [
    {
      "slug": "2016-01-27-alphago",
      "date": "2016-01-27",
      "provider": "google",
      "importance": "major",
      "title": "AlphaGo",
      "headline": "Google launches AlphaGo",
      "detail": "DeepMind's AlphaGo beat European champion Fan Hui 5-0, the first program to beat a professional Go player.",
      "tags": [
        "model-releases"
      ],
      "related": [
        "2016-03-16-alphago-beats-lee-sedol-4-1",
        "2017-10-18-alphago-zero"
      ],
      "source": "https://blog.google/innovation-and-ai/products/alphago-machine-learning-game-go/",
      "models": [
        "alphago"
      ],
      "source_title": "AlphaGo: using machine learning to master the ancient game of Go",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2016-01-27-alphago",
      "image": "https://frontierlog.ai/og/2016-01-27-alphago.png"
    },
    {
      "slug": "2016-09-08-wavenet",
      "date": "2016-09-08",
      "provider": "google",
      "importance": "major",
      "title": "WaveNet",
      "headline": "Google launches WaveNet",
      "detail": "DeepMind's generative model of raw audio produces more natural speech than the best existing text-to-speech systems.",
      "tags": [
        "model-releases",
        "voice"
      ],
      "related": [],
      "source": "https://deepmind.google/blog/wavenet-a-generative-model-for-raw-audio/",
      "models": [
        "wavenet"
      ],
      "source_title": "WaveNet: A generative model for raw audio",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2016-09-08-wavenet",
      "image": "https://frontierlog.ai/og/2016-09-08-wavenet.png"
    },
    {
      "slug": "2016-09-27-google-neural-machine-translation",
      "date": "2016-09-27",
      "provider": "google",
      "importance": "major",
      "title": "Google Neural Machine Translation",
      "headline": "Google launches Neural Machine Translation",
      "detail": "Google Translate now uses GNMT for all Chinese-to-English translations, cutting translation errors by 55%-85%.",
      "tags": [
        "model-releases"
      ],
      "related": [],
      "source": "https://research.google/blog/a-neural-network-for-machine-translation-at-production-scale/",
      "models": [
        "google-neural-machine-translation"
      ],
      "source_title": "A Neural Network for Machine Translation, at Production Scale",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2016-09-27-google-neural-machine-translation",
      "image": "https://frontierlog.ai/og/2016-09-27-google-neural-machine-translation.png"
    },
    {
      "slug": "2017-10-18-alphago-zero",
      "date": "2017-10-18",
      "provider": "google",
      "importance": "major",
      "title": "AlphaGo Zero",
      "headline": "Google launches AlphaGo Zero",
      "detail": "DeepMind's AlphaGo Zero learns Go by self-play alone and beats the champion-defeating AlphaGo 100-0.",
      "tags": [
        "model-releases"
      ],
      "related": [
        "2016-01-27-alphago",
        "2018-12-06-alphazero"
      ],
      "source": "https://deepmind.google/blog/alphago-zero-starting-from-scratch/",
      "models": [
        "alphago-zero"
      ],
      "source_title": "AlphaGo Zero: Starting from scratch",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2017-10-18-alphago-zero",
      "image": "https://frontierlog.ai/og/2017-10-18-alphago-zero.png"
    },
    {
      "slug": "2018-06-11-gpt-1",
      "date": "2018-06-11",
      "provider": "openai",
      "importance": "major",
      "title": "GPT-1",
      "headline": "OpenAI launches GPT-1",
      "detail": "Transformers plus unsupervised pre-training set state-of-the-art results across language tasks; system released.",
      "tags": [
        "model-releases"
      ],
      "related": [
        "2019-02-14-gpt-2"
      ],
      "source": "https://openai.com/index/language-unsupervised",
      "models": [
        "gpt-1"
      ],
      "source_title": "Improving language understanding with unsupervised learning",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2018-06-11-gpt-1",
      "image": "https://frontierlog.ai/og/2018-06-11-gpt-1.png"
    },
    {
      "slug": "2018-11-02-bert",
      "date": "2018-11-02",
      "provider": "google",
      "importance": "milestone",
      "title": "BERT",
      "headline": "Google releases BERT",
      "detail": "Google open-sources BERT, a bidirectional pre-trained language model, state of the art on 11 NLP tasks.",
      "tags": [
        "model-releases",
        "open-weights"
      ],
      "related": [],
      "source": "https://research.google/blog/open-sourcing-bert-state-of-the-art-pre-training-for-natural-language-processing/",
      "models": [
        "bert"
      ],
      "source_title": "Open Sourcing BERT: State-of-the-Art Pre-training for Natural Language Processin",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2018-11-02-bert",
      "image": "https://frontierlog.ai/og/2018-11-02-bert.png"
    },
    {
      "slug": "2018-12-06-alphazero",
      "date": "2018-12-06",
      "provider": "google",
      "importance": "major",
      "title": "AlphaZero",
      "headline": "Google launches AlphaZero",
      "detail": "DeepMind's AlphaZero teaches itself chess, shogi and Go, beating a world-champion program in each.",
      "tags": [
        "model-releases"
      ],
      "related": [
        "2017-10-18-alphago-zero"
      ],
      "source": "https://deepmind.google/blog/alphazero-shedding-new-light-on-chess-shogi-and-go/",
      "models": [
        "alphazero"
      ],
      "source_title": "AlphaZero: Shedding new light on chess, shogi, and Go",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2018-12-06-alphazero",
      "image": "https://frontierlog.ai/og/2018-12-06-alphazero.png"
    },
    {
      "slug": "2019-02-14-gpt-2",
      "date": "2019-02-14",
      "provider": "openai",
      "importance": "major",
      "title": "GPT-2",
      "headline": "OpenAI launches GPT-2",
      "detail": "A 1.5B-parameter language model trained to predict the next word; only a much smaller model was released.",
      "tags": [
        "model-releases"
      ],
      "related": [
        "2018-06-11-gpt-1",
        "2020-05-28-gpt-3"
      ],
      "source": "https://openai.com/index/better-language-models",
      "models": [
        "gpt-2"
      ],
      "source_title": "Better language models and their implications",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2019-02-14-gpt-2",
      "image": "https://frontierlog.ai/og/2019-02-14-gpt-2.png"
    },
    {
      "slug": "2020-02-24-t5-text-to-text-transformer",
      "date": "2020-02-24",
      "provider": "google",
      "importance": "major",
      "title": "T5: text-to-text Transformer",
      "headline": "Google releases T5, the text-to-text Transformer",
      "detail": "Google Research open-sources T5, an 11B-parameter text-to-text model, with the C4 dataset.",
      "tags": [
        "model-releases",
        "open-weights"
      ],
      "related": [],
      "source": "https://research.google/blog/exploring-transfer-learning-with-t5-the-text-to-text-transfer-transformer/",
      "models": [
        "t5"
      ],
      "source_title": "Exploring Transfer Learning with T5: the Text-To-Text Transfer Transformer",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2020-02-24-t5-text-to-text-transformer",
      "image": "https://frontierlog.ai/og/2020-02-24-t5-text-to-text-transformer.png"
    },
    {
      "slug": "2020-05-28-gpt-3",
      "date": "2020-05-28",
      "provider": "openai",
      "importance": "milestone",
      "title": "GPT-3",
      "headline": "OpenAI launches GPT-3",
      "detail": "A 175-billion-parameter language model that performs new tasks from a few examples, without fine-tuning.",
      "tags": [
        "model-releases"
      ],
      "related": [
        "2019-02-14-gpt-2",
        "2023-03-14-gpt-4",
        "2020-06-11-openai-api",
        "2022-01-27-instructgpt"
      ],
      "source": "https://openai.com/index/language-models-are-few-shot-learners",
      "models": [
        "gpt-3"
      ],
      "source_title": "Language models are few-shot learners",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2020-05-28-gpt-3",
      "image": "https://frontierlog.ai/og/2020-05-28-gpt-3.png"
    },
    {
      "slug": "2020-11-30-alphafold-2-protein-folding-solved",
      "date": "2020-11-30",
      "provider": "google",
      "importance": "milestone",
      "title": "AlphaFold 2: protein folding solved",
      "headline": "Google's AlphaFold 2 solves protein folding",
      "detail": "DeepMind's latest AlphaFold scores a median 92.4 GDT at CASP14, recognised as a solution to protein folding.",
      "tags": [
        "model-releases",
        "science"
      ],
      "related": [
        "2021-07-22-alphafold-open-sourced-and-alphafold-db",
        "2024-10-09-nobel-prize-in-chemistry-for-alphafold"
      ],
      "source": "https://deepmind.google/blog/alphafold-a-solution-to-a-50-year-old-grand-challenge-in-biology/",
      "models": [
        "alphafold-2"
      ],
      "source_title": "AlphaFold: a solution to a 50-year-old grand challenge in biology",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2020-11-30-alphafold-2-protein-folding-solved",
      "image": "https://frontierlog.ai/og/2020-11-30-alphafold-2-protein-folding-solved.png"
    },
    {
      "slug": "2020-12-03-vision-transformer-vit",
      "date": "2020-12-03",
      "provider": "google",
      "importance": "major",
      "title": "Vision Transformer (ViT)",
      "headline": "Google releases Vision Transformer (ViT)",
      "detail": "Google Research's ViT applies the Transformer to image patches, beating comparable CNNs with four times less compute.",
      "tags": [
        "model-releases",
        "vision",
        "open-weights"
      ],
      "related": [],
      "source": "https://research.google/blog/transformers-for-image-recognition-at-scale/",
      "models": [
        "vit"
      ],
      "source_title": "Transformers for Image Recognition at Scale",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2020-12-03-vision-transformer-vit",
      "image": "https://frontierlog.ai/og/2020-12-03-vision-transformer-vit.png"
    },
    {
      "slug": "2021-01-05-clip",
      "date": "2021-01-05",
      "provider": "openai",
      "importance": "major",
      "title": "CLIP",
      "headline": "OpenAI launches CLIP",
      "detail": "A neural network that learns visual concepts from natural language and classifies images zero-shot.",
      "tags": [
        "model-releases",
        "vision"
      ],
      "related": [],
      "source": "https://openai.com/index/clip",
      "models": [
        "clip"
      ],
      "source_title": "CLIP: Connecting text and images",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2021-01-05-clip",
      "image": "https://frontierlog.ai/og/2021-01-05-clip.png"
    },
    {
      "slug": "2021-01-05-dall-e",
      "date": "2021-01-05",
      "provider": "openai",
      "importance": "major",
      "title": "DALL·E",
      "headline": "OpenAI launches DALL·E",
      "detail": "A 12-billion parameter version of GPT-3 that generates images from text captions.",
      "tags": [
        "model-releases",
        "image-generation"
      ],
      "related": [
        "2022-07-20-dall-e-available-in-beta",
        "2023-10-19-dall-e-3-in-chatgpt"
      ],
      "source": "https://openai.com/index/dall-e",
      "models": [
        "dall-e"
      ],
      "source_title": "DALL·E: Creating images from text",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2021-01-05-dall-e",
      "image": "https://frontierlog.ai/og/2021-01-05-dall-e.png"
    },
    {
      "slug": "2021-05-18-lamda",
      "date": "2021-05-18",
      "provider": "google",
      "importance": "major",
      "title": "LaMDA",
      "headline": "Google launches LaMDA",
      "detail": "Google's conversational language model, trained on dialogue, can talk freely about a seemingly endless number of topics.",
      "tags": [
        "model-releases"
      ],
      "related": [],
      "source": "https://blog.google/innovation-and-ai/products/lamda/",
      "models": [
        "lamda"
      ],
      "source_title": "LaMDA: our breakthrough conversation technology",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2021-05-18-lamda",
      "image": "https://frontierlog.ai/og/2021-05-18-lamda.png"
    },
    {
      "slug": "2021-08-10-codex",
      "date": "2021-08-10",
      "provider": "openai",
      "importance": "major",
      "title": "Codex",
      "headline": "OpenAI launches Codex",
      "detail": "Codex, the model behind GitHub Copilot, translates natural language to code; private beta in the API.",
      "tags": [
        "model-releases",
        "coding"
      ],
      "related": [
        "2025-05-16-codex"
      ],
      "source": "https://openai.com/index/openai-codex",
      "models": [
        "openai-codex"
      ],
      "source_title": "OpenAI Codex",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2021-08-10-codex",
      "image": "https://frontierlog.ai/og/2021-08-10-codex.png"
    },
    {
      "slug": "2021-12-09-glam",
      "date": "2021-12-09",
      "provider": "google",
      "importance": "major",
      "title": "GLaM",
      "headline": "Google launches GLaM",
      "detail": "Google Brain's trillion-weight sparse mixture-of-experts (MoE) model; competitive with GPT-3, more efficiently.",
      "tags": [
        "model-releases",
        "moe"
      ],
      "related": [],
      "source": "https://research.google/blog/more-efficient-in-context-learning-with-glam/",
      "models": [
        "glam"
      ],
      "source_title": "More Efficient In-Context Learning with GLaM",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2021-12-09-glam",
      "image": "https://frontierlog.ai/og/2021-12-09-glam.png"
    },
    {
      "slug": "2022-01-27-instructgpt",
      "date": "2022-01-27",
      "provider": "openai",
      "importance": "major",
      "title": "InstructGPT",
      "headline": "OpenAI launches InstructGPT",
      "detail": "Models trained with humans in the loop follow instructions better than GPT-3; now the API default.",
      "tags": [
        "model-releases",
        "safety"
      ],
      "related": [
        "2020-05-28-gpt-3",
        "2017-06-13-learning-from-human-preferences"
      ],
      "source": "https://openai.com/index/instruction-following",
      "models": [
        "instructgpt"
      ],
      "source_title": "Aligning language models to follow instructions",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2022-01-27-instructgpt",
      "image": "https://frontierlog.ai/og/2022-01-27-instructgpt.png"
    },
    {
      "slug": "2022-04-04-palm",
      "date": "2022-04-04",
      "provider": "google",
      "importance": "major",
      "title": "PaLM",
      "headline": "Google launches PaLM",
      "detail": "Google's 540B-parameter model, trained with Pathways on 6,144 TPU chips, sets few-shot records.",
      "tags": [
        "model-releases"
      ],
      "related": [
        "2023-05-10-palm-2"
      ],
      "source": "https://research.google/blog/pathways-language-model-palm-scaling-to-540-billion-parameters-for-breakthrough-performance/",
      "models": [
        "palm"
      ],
      "source_title": "Pathways Language Model (PaLM): Scaling to 540 Billion Parameters for Breakthrou",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2022-04-04-palm",
      "image": "https://frontierlog.ai/og/2022-04-04-palm.png"
    },
    {
      "slug": "2022-05-12-gato-a-generalist-agent",
      "date": "2022-05-12",
      "provider": "google",
      "importance": "major",
      "title": "Gato: a generalist agent",
      "headline": "Google introduces Gato, a generalist agent",
      "detail": "DeepMind's Gato uses one network to play Atari, caption images, chat and stack blocks with a robot arm.",
      "tags": [
        "model-releases"
      ],
      "related": [],
      "source": "https://deepmind.google/blog/a-generalist-agent/",
      "models": [
        "gato"
      ],
      "source_title": "A Generalist Agent",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2022-05-12-gato-a-generalist-agent",
      "image": "https://frontierlog.ai/og/2022-05-12-gato-a-generalist-agent.png"
    },
    {
      "slug": "2022-09-21-whisper",
      "date": "2022-09-21",
      "provider": "openai",
      "importance": "major",
      "title": "Whisper",
      "headline": "OpenAI releases Whisper",
      "detail": "An open-sourced speech recognition model approaching human-level robustness and accuracy on English.",
      "tags": [
        "model-releases",
        "voice",
        "open-weights"
      ],
      "related": [],
      "source": "https://openai.com/index/whisper",
      "models": [
        "whisper"
      ],
      "source_title": "Introducing Whisper",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2022-09-21-whisper",
      "image": "https://frontierlog.ai/og/2022-09-21-whisper.png"
    },
    {
      "slug": "2022-12-08-alphacode",
      "date": "2022-12-08",
      "provider": "google",
      "importance": "minor",
      "title": "AlphaCode",
      "headline": "Google launches AlphaCode",
      "detail": "DeepMind's AlphaCode, now published in Science, writes programs at a competitive level in coding contests.",
      "tags": [
        "model-releases",
        "coding"
      ],
      "related": [],
      "source": "https://deepmind.google/blog/competitive-programming-with-alphacode/",
      "models": [
        "alphacode"
      ],
      "source_title": "Competitive programming with AlphaCode",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2022-12-08-alphacode",
      "image": "https://frontierlog.ai/og/2022-12-08-alphacode.png"
    },
    {
      "slug": "2022-12-13-rt-1-robotics-transformer",
      "date": "2022-12-13",
      "provider": "google",
      "importance": "minor",
      "title": "RT-1 Robotics Transformer",
      "headline": "Google launches RT-1 Robotics Transformer",
      "detail": "Google Research's multi-task robot model, trained on 130k episodes over 700+ tasks; its code is open source.",
      "tags": [
        "model-releases"
      ],
      "related": [
        "2023-07-28-rt-2"
      ],
      "source": "https://research.google/blog/rt-1-robotics-transformer-for-real-world-control-at-scale/",
      "models": [
        "rt-1"
      ],
      "source_title": "RT-1: Robotics Transformer for real-world control at scale",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2022-12-13-rt-1-robotics-transformer",
      "image": "https://frontierlog.ai/og/2022-12-13-rt-1-robotics-transformer.png"
    },
    {
      "slug": "2023-02-24-llama",
      "date": "2023-02-24",
      "provider": "meta",
      "importance": "milestone",
      "title": "LLaMA",
      "headline": "Meta releases LLaMA",
      "detail": "Foundation models from 7B to 65B parameters, with weights released to researchers for noncommercial use.",
      "tags": [
        "model-releases",
        "open-weights"
      ],
      "related": [
        "2023-07-18-llama-2"
      ],
      "source": "https://ai.meta.com/blog/large-language-model-llama-meta-ai/",
      "models": [
        "llama"
      ],
      "source_title": "Introducing LLaMA: A foundational, 65-billion-parameter language model",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-02-24-llama",
      "image": "https://frontierlog.ai/og/2023-02-24-llama.png"
    },
    {
      "slug": "2023-03-14-gpt-4",
      "date": "2023-03-14",
      "provider": "openai",
      "importance": "milestone",
      "title": "GPT-4",
      "headline": "OpenAI launches GPT-4",
      "detail": "Multimodal model accepting image and text inputs and emitting text, with human-level performance on various benchmarks.",
      "tags": [
        "model-releases",
        "vision"
      ],
      "related": [
        "2023-11-06-gpt-4-turbo",
        "2024-04-24-gpt-4-api-general-availability",
        "2020-05-28-gpt-3"
      ],
      "source": "https://openai.com/index/gpt-4-research",
      "models": [
        "gpt-4"
      ],
      "source_title": "GPT-4",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-03-14-gpt-4",
      "image": "https://frontierlog.ai/og/2023-03-14-gpt-4.png"
    },
    {
      "slug": "2023-03-14-claude-and-claude-instant",
      "date": "2023-03-14",
      "provider": "anthropic",
      "importance": "milestone",
      "title": "Claude and Claude Instant",
      "headline": "Anthropic launches Claude and Claude Instant",
      "detail": "Anthropic offers Claude and the lighter, faster Claude Instant through a chat interface and API.",
      "tags": [
        "model-releases"
      ],
      "related": [
        "2023-07-11-claude-2"
      ],
      "source": "https://www.anthropic.com/news/introducing-claude",
      "models": [
        "claude",
        "claude-instant"
      ],
      "source_title": "Introducing Claude",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-03-14-claude-and-claude-instant",
      "image": "https://frontierlog.ai/og/2023-03-14-claude-and-claude-instant.png"
    },
    {
      "slug": "2023-05-10-palm-2",
      "date": "2023-05-10",
      "provider": "google",
      "importance": "major",
      "title": "PaLM 2",
      "headline": "Google launches PaLM 2",
      "detail": "Language model with stronger multilingual, reasoning and coding skills, in four sizes, powering 25+ products.",
      "tags": [
        "model-releases"
      ],
      "related": [
        "2022-04-04-palm"
      ],
      "source": "https://blog.google/innovation-and-ai/products/google-palm-2-ai-large-language-model/",
      "models": [
        "palm-2"
      ],
      "source_title": "Introducing PaLM 2",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-05-10-palm-2",
      "image": "https://frontierlog.ai/og/2023-05-10-palm-2.png"
    },
    {
      "slug": "2023-07-11-claude-2",
      "date": "2023-07-11",
      "provider": "anthropic",
      "importance": "major",
      "title": "Claude 2",
      "headline": "Anthropic launches Claude 2",
      "detail": "New model with a 100K token context window, available via API and the claude.ai beta in the US and UK.",
      "tags": [
        "model-releases"
      ],
      "related": [
        "2023-03-14-claude-and-claude-instant",
        "2023-11-21-claude-2-1"
      ],
      "source": "https://www.anthropic.com/news/claude-2",
      "models": [
        "claude-2"
      ],
      "source_title": "Claude 2",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-07-11-claude-2",
      "image": "https://frontierlog.ai/og/2023-07-11-claude-2.png"
    },
    {
      "slug": "2023-07-18-llama-2",
      "date": "2023-07-18",
      "provider": "meta",
      "importance": "milestone",
      "title": "Llama 2",
      "headline": "Meta releases Llama 2",
      "detail": "Pretrained and chat models from 7B to 70B parameters, free for research and commercial use.",
      "tags": [
        "model-releases",
        "open-weights"
      ],
      "related": [
        "2023-02-24-llama",
        "2024-07-23-llama-3-1-405b"
      ],
      "source": "https://about.fb.com/news/2023/07/llama-2/",
      "models": [
        "llama-2"
      ],
      "source_title": "Meta and Microsoft Introduce the Next Generation of Llama",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-07-18-llama-2",
      "image": "https://frontierlog.ai/og/2023-07-18-llama-2.png"
    },
    {
      "slug": "2023-07-28-rt-2",
      "date": "2023-07-28",
      "provider": "google",
      "importance": "major",
      "title": "RT-2",
      "headline": "Google launches RT-2",
      "detail": "A vision-language-action model that learns from web and robot data and turns instructions into robot actions.",
      "tags": [
        "model-releases",
        "vision"
      ],
      "related": [
        "2022-12-13-rt-1-robotics-transformer"
      ],
      "source": "https://deepmind.google/blog/rt-2-new-model-translates-vision-and-language-into-action/",
      "models": [
        "rt-2"
      ],
      "source_title": "RT-2: New model translates vision and language into action",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-07-28-rt-2",
      "image": "https://frontierlog.ai/og/2023-07-28-rt-2.png"
    },
    {
      "slug": "2023-08-09-claude-instant-1-2",
      "date": "2023-08-09",
      "provider": "anthropic",
      "importance": "minor",
      "title": "Claude Instant 1.2",
      "headline": "Anthropic launches Claude Instant 1.2",
      "detail": "A faster, lower-priced model with gains in math, coding, reasoning and safety, available through the API.",
      "tags": [
        "model-releases"
      ],
      "related": [],
      "source": "https://www.anthropic.com/news/releasing-claude-instant-1-2",
      "models": [
        "claude-instant-1-2"
      ],
      "source_title": "Releasing Claude Instant 1.2",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-08-09-claude-instant-1-2",
      "image": "https://frontierlog.ai/og/2023-08-09-claude-instant-1-2.png"
    },
    {
      "slug": "2023-09-19-alphamissense",
      "date": "2023-09-19",
      "provider": "google",
      "importance": "minor",
      "title": "AlphaMissense",
      "headline": "Google launches AlphaMissense",
      "detail": "Google DeepMind classifies the effects of 71 million missense mutations to help pinpoint causes of disease.",
      "tags": [
        "model-releases",
        "health",
        "science"
      ],
      "related": [],
      "source": "https://deepmind.google/blog/a-catalogue-of-genetic-mutations-to-help-pinpoint-the-cause-of-diseases/",
      "models": [
        "alphamissense"
      ],
      "source_title": "A catalogue of genetic mutations to help pinpoint the cause of diseases",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-09-19-alphamissense",
      "image": "https://frontierlog.ai/og/2023-09-19-alphamissense.png"
    },
    {
      "slug": "2023-11-06-gpt-4-turbo",
      "date": "2023-11-06",
      "provider": "openai",
      "importance": "major",
      "title": "GPT-4 Turbo",
      "headline": "OpenAI launches GPT-4 Turbo",
      "detail": "GPT-4 Turbo offers a 128K context window and lower prices, with a Vision version.",
      "tags": [
        "model-releases",
        "long-context"
      ],
      "related": [
        "2023-03-14-gpt-4",
        "2023-11-06-assistants-api"
      ],
      "source": "https://openai.com/index/new-models-and-developer-products-announced-at-devday",
      "models": [
        "gpt-4-turbo",
        "tts",
        "whisper-large-v3"
      ],
      "source_title": "New models and developer products announced at DevDay",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-11-06-gpt-4-turbo",
      "image": "https://frontierlog.ai/og/2023-11-06-gpt-4-turbo.png"
    },
    {
      "slug": "2023-11-14-graphcast",
      "date": "2023-11-14",
      "provider": "google",
      "importance": "major",
      "title": "GraphCast",
      "headline": "Google launches GraphCast",
      "detail": "Google DeepMind's weather model gives 10-day forecasts in under a minute, beating HRES on 90%+ of variables.",
      "tags": [
        "model-releases",
        "science"
      ],
      "related": [],
      "source": "https://deepmind.google/blog/graphcast-ai-model-for-faster-and-more-accurate-global-weather-forecasting/",
      "models": [
        "graphcast"
      ],
      "source_title": "GraphCast: AI model for faster and more accurate global weather forecasting",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-11-14-graphcast",
      "image": "https://frontierlog.ai/og/2023-11-14-graphcast.png"
    },
    {
      "slug": "2023-11-16-lyria",
      "date": "2023-11-16",
      "provider": "google",
      "importance": "minor",
      "title": "Lyria",
      "headline": "Google launches Lyria",
      "detail": "Google DeepMind's music generation model, tested in YouTube Shorts via Dream Track and in Music AI tools.",
      "tags": [
        "model-releases",
        "voice"
      ],
      "related": [],
      "source": "https://deepmind.google/blog/transforming-the-future-of-music-creation/",
      "models": [
        "lyria"
      ],
      "source_title": "Transforming the future of music creation",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-11-16-lyria",
      "image": "https://frontierlog.ai/og/2023-11-16-lyria.png"
    },
    {
      "slug": "2023-11-21-claude-2-1",
      "date": "2023-11-21",
      "provider": "anthropic",
      "importance": "major",
      "title": "Claude 2.1",
      "headline": "Anthropic launches Claude 2.1",
      "detail": "Adds a 200K token context window, 2x fewer false statements than Claude 2.0, system prompts and beta tool use.",
      "tags": [
        "model-releases",
        "tool-use",
        "long-context"
      ],
      "related": [
        "2023-07-11-claude-2",
        "2024-03-04-claude-3-family"
      ],
      "source": "https://www.anthropic.com/news/claude-2-1",
      "models": [
        "claude-2-1"
      ],
      "source_title": "Introducing Claude 2.1",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-11-21-claude-2-1",
      "image": "https://frontierlog.ai/og/2023-11-21-claude-2-1.png"
    },
    {
      "slug": "2023-11-29-gnome-materials-discovery",
      "date": "2023-11-29",
      "provider": "google",
      "importance": "minor",
      "title": "GNoME materials discovery",
      "headline": "Google launches GNoME for materials discovery",
      "detail": "A deep learning tool predicted 2.2 million new crystals; 380,000 stable ones are released to researchers.",
      "tags": [
        "model-releases",
        "science"
      ],
      "related": [],
      "source": "https://deepmind.google/blog/millions-of-new-materials-discovered-with-deep-learning/",
      "models": [
        "gnome"
      ],
      "source_title": "Millions of new materials discovered with deep learning",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-11-29-gnome-materials-discovery",
      "image": "https://frontierlog.ai/og/2023-11-29-gnome-materials-discovery.png"
    },
    {
      "slug": "2023-12-06-gemini-1-0-ultra-pro-and-nano",
      "date": "2023-12-06",
      "provider": "google",
      "importance": "milestone",
      "title": "Gemini 1.0 (Ultra, Pro and Nano)",
      "headline": "Google launches Gemini 1.0 (Ultra, Pro and Nano)",
      "detail": "Google's first natively multimodal models, in three sizes; Ultra leads on 30 of 32 benchmarks at launch.",
      "tags": [
        "model-releases",
        "vision"
      ],
      "related": [
        "2023-12-13-gemini-pro-in-the-gemini-api",
        "2024-02-15-gemini-1-5",
        "2023-12-06-gemini-pro-in-bard"
      ],
      "source": "https://blog.google/innovation-and-ai/technology/ai/google-gemini-ai/",
      "models": [
        "gemini-1-0-nano",
        "gemini-1-0-pro",
        "gemini-1-0-ultra"
      ],
      "source_title": "Introducing Gemini: our largest and most capable AI model",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-12-06-gemini-1-0-ultra-pro-and-nano",
      "image": "https://frontierlog.ai/og/2023-12-06-gemini-1-0-ultra-pro-and-nano.png"
    },
    {
      "slug": "2023-12-11-mixtral-8x7b",
      "date": "2023-12-11",
      "provider": "mistral",
      "importance": "major",
      "title": "Mixtral 8x7B",
      "headline": "Mistral releases Mixtral 8x7B",
      "detail": "Sparse mixture-of-experts (MoE) model with open weights under Apache 2.0; beats Llama 2 70B on most benchmarks.",
      "tags": [
        "model-releases",
        "moe",
        "open-weights"
      ],
      "related": [],
      "source": "https://mistral.ai/news/mixtral-of-experts",
      "models": [
        "mixtral-8x7b"
      ],
      "source_title": "Mixtral of experts",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2023-12-11-mixtral-8x7b",
      "image": "https://frontierlog.ai/og/2023-12-11-mixtral-8x7b.png"
    },
    {
      "slug": "2024-01-17-alphageometry",
      "date": "2024-01-17",
      "provider": "google",
      "importance": "major",
      "title": "AlphaGeometry",
      "headline": "Google launches AlphaGeometry",
      "detail": "DeepMind's neuro-symbolic system solved 25 of 30 Olympiad geometry problems; code and model open-sourced.",
      "tags": [
        "model-releases",
        "science"
      ],
      "related": [],
      "source": "https://deepmind.google/blog/alphageometry-an-olympiad-level-ai-system-for-geometry/",
      "models": [
        "alphageometry"
      ],
      "source_title": "AlphaGeometry: An Olympiad-level AI system for geometry",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2024-01-17-alphageometry",
      "image": "https://frontierlog.ai/og/2024-01-17-alphageometry.png"
    },
    {
      "slug": "2024-02-01-imagen-2",
      "date": "2024-02-01",
      "provider": "google",
      "importance": "minor",
      "title": "Imagen 2",
      "headline": "Google launches Imagen 2",
      "detail": "Google's upgraded text-to-image model rolls out in Bard, ImageFX, Search and Vertex AI, with SynthID.",
      "tags": [
        "model-releases",
        "image-generation"
      ],
      "related": [
        "2024-05-14-imagen-3"
      ],
      "source": "https://blog.google/innovation-and-ai/products/google-imagen-2/",
      "models": [
        "imagen-2"
      ],
      "source_title": "New and better ways to create images with Imagen 2",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2024-02-01-imagen-2",
      "image": "https://frontierlog.ai/og/2024-02-01-imagen-2.png"
    },
    {
      "slug": "2024-02-15-gemini-1-5",
      "date": "2024-02-15",
      "provider": "google",
      "importance": "milestone",
      "title": "Gemini 1.5",
      "headline": "Google launches Gemini 1.5",
      "detail": "Gemini 1.5 Pro, a mixture-of-experts (MoE) model, matches 1.0 Ultra with less compute and offers up to 1M tokens.",
      "tags": [
        "model-releases",
        "long-context",
        "moe"
      ],
      "related": [
        "2023-12-06-gemini-1-0-ultra-pro-and-nano",
        "2024-05-14-gemini-1-5-flash",
        "2024-12-11-gemini-2-0-flash"
      ],
      "source": "https://blog.google/innovation-and-ai/products/google-gemini-next-generation-model-february-2024/",
      "models": [
        "gemini-1-5-pro"
      ],
      "source_title": "Our next-generation model: Gemini 1.5",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2024-02-15-gemini-1-5",
      "image": "https://frontierlog.ai/og/2024-02-15-gemini-1-5.png"
    },
    {
      "slug": "2024-02-15-sora-announced",
      "date": "2024-02-15",
      "provider": "openai",
      "importance": "major",
      "title": "Sora announced",
      "headline": "OpenAI announces Sora",
      "detail": "A text-conditional diffusion model trained jointly on videos and images to generate video.",
      "tags": [
        "model-releases",
        "video-generation"
      ],
      "related": [
        "2024-12-09-sora-is-here"
      ],
      "source": "https://openai.com/index/video-generation-models-as-world-simulators",
      "models": [
        "sora"
      ],
      "source_title": "Video generation models as world simulators",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2024-02-15-sora-announced",
      "image": "https://frontierlog.ai/og/2024-02-15-sora-announced.png"
    },
    {
      "slug": "2024-02-21-gemma",
      "date": "2024-02-21",
      "provider": "google",
      "importance": "major",
      "title": "Gemma",
      "headline": "Google releases Gemma",
      "detail": "Open-weights Gemma 2B and 7B models, built from Gemini research, with a Responsible Generative AI Toolkit.",
      "tags": [
        "model-releases",
        "open-weights"
      ],
      "related": [],
      "source": "https://blog.google/innovation-and-ai/technology/developers-tools/gemma-open-models/",
      "models": [
        "gemma"
      ],
      "source_title": "Gemma: Introducing new state-of-the-art open models",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2024-02-21-gemma",
      "image": "https://frontierlog.ai/og/2024-02-21-gemma.png"
    },
    {
      "slug": "2024-03-04-claude-3-family",
      "date": "2024-03-04",
      "provider": "anthropic",
      "importance": "milestone",
      "title": "Claude 3 family",
      "headline": "Anthropic launches the Claude 3 family",
      "detail": "Haiku, Sonnet and Opus models, in ascending order of capability, set new benchmarks at launch.",
      "tags": [
        "model-releases",
        "vision"
      ],
      "related": [
        "2024-03-13-claude-3-haiku",
        "2023-11-21-claude-2-1"
      ],
      "source": "https://www.anthropic.com/news/claude-3-family",
      "models": [
        "claude-3-opus",
        "claude-3-haiku",
        "claude-3-sonnet"
      ],
      "source_title": "Introducing the next generation of Claude",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2024-03-04-claude-3-family",
      "image": "https://frontierlog.ai/og/2024-03-04-claude-3-family.png"
    },
    {
      "slug": "2024-03-13-claude-3-haiku",
      "date": "2024-03-13",
      "provider": "anthropic",
      "importance": "minor",
      "title": "Anthropic releases Claude 3 Haiku",
      "headline": "Anthropic releases Claude 3 Haiku",
      "detail": "The fastest, most affordable model in its intelligence class, in the API and for Claude Pro users.",
      "tags": [
        "model-releases",
        "small-models"
      ],
      "related": [
        "2024-03-04-claude-3-family"
      ],
      "source": "https://www.anthropic.com/news/claude-3-haiku",
      "models": [
        "claude-3-haiku"
      ],
      "source_title": "Claude 3 Haiku: our fastest model yet",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2024-03-13-claude-3-haiku",
      "image": "https://frontierlog.ai/og/2024-03-13-claude-3-haiku.png"
    },
    {
      "slug": "2024-03-17-grok-1",
      "date": "2024-03-17",
      "provider": "xai",
      "importance": "major",
      "title": "Grok-1",
      "headline": "xAI releases Grok-1",
      "detail": "Open release of the base model weights and architecture of a 314B-parameter mixture-of-experts (MoE) model, under the Apache 2.0 license.",
      "tags": [
        "model-releases",
        "open-weights"
      ],
      "related": [],
      "source": "https://x.ai/news/grok-os",
      "models": [
        "grok-1"
      ],
      "source_title": "Open Release of Grok-1",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2024-03-17-grok-1",
      "image": "https://frontierlog.ai/og/2024-03-17-grok-1.png"
    },
    {
      "slug": "2024-04-24-snowflake-arctic",
      "date": "2024-04-24",
      "provider": "snowflake",
      "importance": "major",
      "title": "Snowflake Arctic",
      "headline": "Snowflake releases Snowflake Arctic",
      "detail": "Open mixture-of-experts (MoE) model with 480B total and 17B active parameters, built for enterprise tasks such as SQL and coding, under Apache 2.0.",
      "tags": [
        "model-releases",
        "open-weights"
      ],
      "related": [],
      "source": "https://www.snowflake.com/en/blog/arctic-open-efficient-foundation-language-models-snowflake/",
      "models": [
        "snowflake-arctic"
      ],
      "source_title": "Snowflake Arctic: The Best LLM for Enterprise AI — Efficiently Intelligent, Truly Open",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2024-04-24-snowflake-arctic",
      "image": "https://frontierlog.ai/og/2024-04-24-snowflake-arctic.png"
    },
    {
      "slug": "2024-05-08-alphafold-3",
      "date": "2024-05-08",
      "provider": "google",
      "importance": "major",
      "title": "AlphaFold 3",
      "headline": "Google launches AlphaFold 3",
      "detail": "Predicts structures and interactions of proteins, DNA, RNA and ligands; free via the new AlphaFold Server.",
      "tags": [
        "model-releases",
        "science"
      ],
      "related": [],
      "source": "https://blog.google/innovation-and-ai/products/google-deepmind-isomorphic-alphafold-3-ai-model/",
      "models": [
        "alphafold-3"
      ],
      "source_title": "AlphaFold 3 predicts the structure and interactions of all of life’s molecules",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2024-05-08-alphafold-3",
      "image": "https://frontierlog.ai/og/2024-05-08-alphafold-3.png"
    },
    {
      "slug": "2024-05-13-gpt-4o",
      "date": "2024-05-13",
      "provider": "openai",
      "importance": "milestone",
      "title": "GPT-4o",
      "headline": "OpenAI launches GPT-4o",
      "detail": "New flagship model that reasons across audio, vision and text in real time.",
      "tags": [
        "model-releases",
        "voice"
      ],
      "related": [
        "2024-07-18-gpt-4o-mini",
        "2025-03-25-4o-image-generation",
        "2025-04-29-gpt-4o-sycophancy-rollback",
        "2026-01-29-retire-gpt-4o"
      ],
      "source": "https://openai.com/index/hello-gpt-4o",
      "models": [
        "gpt-4o"
      ],
      "source_title": "Hello GPT-4o",
      "source_archived": null,
      "source_http_status": null,
      "followups": [],
      "revealed_as": null,
      "stealth_previews": [],
      "reveal_source": null,
      "url": "https://frontierlog.ai/event/2024-05-13-gpt-4o",
      "image": "https://frontierlog.ai/og/2024-05-13-gpt-4o.png"
    }
  ],
  "next_cursor": "WzUwLCAiNjhhYWQ2NWJhNDI5Il0"
}