Frontier Log

Claude Opus 5.5 vs Gemini 4 Neon

Claude Opus 5.5Gemini 4 Neon
Released
LineupFable › Opus › Sonnet › HaikuCarbon = Argon › Neon › Hydrogen
Input / 1M tokens$4$3
Cached input / 1M tokens$0.20$0.15
Output / 1M tokens$20$15
Context window1M1M
Output limit128K65K
Benchmarks
Terminal-Bench Science 0.158.7%170.5%2
FrontierCode 2.0 Score, all tasks (132)45.5%346.2%4
FrontierCode 2.0 Pass rate, all tasks (132)51.8%5Claude Code50.8%6Antigravity CLI
Chartography No tools64.4%773.5%8

Download image

Claude Opus 5.5Gemini 4 Neon
Other benchmarks
Humanity's Last Exam No tools, full set (2,500 questions)64.4%No data
Humanity's Last Exam With tools, full set (2,500 questions)67.7%9No data
HLE-Diamond No toolsNo data71.3%10
SWE-bench Pro Set not specified89.9%11No data
Terminal-Bench 4.066.4%12No data
Terminal-Bench 4.165.8%13No data
AutomationBench 1.0.6 Private set40.0%14No data
AutomationBench, release not specified Private setNo data56.3%15
OSWorld 2.1 Full set (108 tasks), partial score81.8%16No data
OSWorld 2.1 Full set (108 tasks), strict pass rate48.7%17No data
OSWorld 2.1 Offline set (82 tasks), partial score87.2%18No data
OSWorld 2.1 Offline set (82 tasks), strict pass rate53.2%19No data
OSWorld 2.0, release not specified Anthropic's Sep 10 files, partial81.8%20No data
OSWorld 2.0, release not specified Anthropic's Sep 10 files, strict48.7%21No data
FrontierCode 1.1 Main set54.6%22No data
FrontierCode 1.1 Extended set65.3%23No data
FrontierSWE v262.3%24No data
DeepSWE v1.174.2%25No data
Chartography With tools89.0%26No data
SEC-bench Pro 260505No data90.5%27
Agents' Last Exam v1.1 Score, all tasks (122)No data75.8%28
GraphWalks BFS, 256K + 1M problemsNo data95.2%29
ExploitGym Mitigations off, scoreNo data48.3%30
ExploitGym Mitigations off, successful exploits300 of 86931No data

Notes

  1. Terminal-Bench Science 0.1, Claude Opus 5.5: max effort · standard error ±3.5 to 5 points
  2. Terminal-Bench Science 0.1, Gemini 4 Neon: mini-swe-agent vision harness · highest thinking · average of 5 attempts per task
  3. FrontierCode 2.0, Score, all tasks (132), Claude Opus 5.5: Run by Cognition · Claude Code, per the system card · highest-available effort, max effort in the post's FrontierCode 2.0 chart · mean@5, per the system card · Scored before Cognition updated its cheating scanner to remove false positives and rescored all results on 11 Nov 2026
  4. FrontierCode 2.0, Score, all tasks (132), Gemini 4 Neon: Run by Cognition · high thinking · set and measure of the official public leaderboard the evaluation methodology cites
  5. FrontierCode 2.0, Pass rate, all tasks (132), Claude Opus 5.5: Cognition’s leaderboard, 18 Nov 2026 · Claude Code · max effort
  6. FrontierCode 2.0, Pass rate, all tasks (132), Gemini 4 Neon: Cognition’s leaderboard, 18 Nov 2026 · Antigravity CLI · high effort
  7. Chartography, No tools, Claude Opus 5.5: max effort · adaptive thinking · average of 5 runs · Gemini 3.5 Flash grader
  8. Chartography, No tools, Gemini 4 Neon: highest thinking
  9. Humanity's Last Exam, With tools, full set (2,500 questions), Claude Opus 5.5: max effort
  10. HLE-Diamond, No tools, Gemini 4 Neon: Run by Artificial Analysis · GPT-5.6 Luna (medium) judge with the official HLE judge prompt, per Artificial Analysis’s methodology
  11. SWE-bench Pro, Set not specified, Claude Opus 5.5: max effort · adaptive thinking · average of 5 trials
  12. Terminal-Bench 4.0, Claude Opus 5.5: xhigh effort · standard error ±2.6 points
  13. Terminal-Bench 4.1, Claude Opus 5.5: Stanford University and Laude Institute’s leaderboard, 18 Nov 2026 · Claude Code · max effort
  14. AutomationBench 1.0.6, Private set, Claude Opus 5.5: Run by Zapier · without fallback models · set and release 1.0.6 named in the Claude Sonnet 5.5 system card, which prints the same figure
  15. AutomationBench, release not specified, Private set, Gemini 4 Neon: Run by Zapier · set of the official public leaderboard the evaluation methodology cites
  16. OSWorld 2.1, Full set (108 tasks), partial score, Claude Opus 5.5: max effort · mean of 5 runs · September 10, 2026 task files · the system card labels it OSWorld 2.0
  17. OSWorld 2.1, Full set (108 tasks), strict pass rate, Claude Opus 5.5: max effort · mean of 5 runs · September 10, 2026 task files
  18. OSWorld 2.1, Offline set (82 tasks), partial score, Claude Opus 5.5: max effort · mean of 5 attempts
  19. OSWorld 2.1, Offline set (82 tasks), strict pass rate, Claude Opus 5.5: max effort · mean of 5 attempts
  20. OSWorld 2.0, release not specified, Anthropic's Sep 10 files, partial, Claude Opus 5.5: max effort · mean of 5 runs
  21. OSWorld 2.0, release not specified, Anthropic's Sep 10 files, strict, Claude Opus 5.5: max effort · mean of 5 runs
  22. FrontierCode 1.1, Main set, Claude Opus 5.5: Run by Cognition · Claude Code harness · medium effort
  23. FrontierCode 1.1, Extended set, Claude Opus 5.5: Run by Cognition · Claude Code harness · medium effort
  24. FrontierSWE v2, Claude Opus 5.5: Run by Proximal · Proximal's agent harness · highest available effort · mean of 5 trials per task, per the system card · 20-hour budget · safeguards on
  25. DeepSWE v1.1, Claude Opus 5.5: average of 5 trials
  26. Chartography, With tools, Claude Opus 5.5: max effort · adaptive thinking
  27. SEC-bench Pro 260505, Gemini 4 Neon: Antigravity harness · pass@1, mean of 3 attempts · 183-task JavaScript engine split (May 2026) · LLM judge for crashes in both builds · safety filters off
  28. Agents' Last Exam v1.1, Score, all tasks (122), Gemini 4 Neon: Antigravity-CLI version 1.4.8 harness · highest thinking · 5-hour window · safety filters on
  29. GraphWalks, BFS, 256K + 1M problems, Gemini 4 Neon: highest thinking · 200 problems between 256K and 1M tokens
  30. ExploitGym, Mitigations off, score, Gemini 4 Neon: Antigravity harness · single attempt per task · no internet access · 10h time cap per task · safety filters off · mitigations off (the default setting), per the model evaluation
  31. ExploitGym, Mitigations off, successful exploits, Claude Opus 5.5: 6h wall-clock budget · agent judge confirms the intended vulnerability

Related

All comparisons