GPT-5.5
OpenAI·Released Apr 23, 2026
Updated Oct 8, 9:18 PM ET
Overall score
59.0
Rank #21 overall
Evidence
30 results
85% of scored kinds covered
Context window
1.1M tokens
From OpenRouter
API price, USD per 1M tokens
$5 / $30
Input / output · cached input $0.50
The score is on the board's scale (reference average 50, about 15 points per standard deviation), not a percentage.
By domain
Each domain is scored on the same scale as the overall score. Coverage is the share of the domain's kinds of evaluation with a result; it says how much evidence there is, not how good the model is.
Every result behind the score
The result as the evaluator publishes it, and the reading it gives on the board's scale. A hard evaluation can give a high reading for a modest result; results near 95% or 5% only say "at least" or "at most".
Coding
7 results · 57.9- SciCodeArtificial Analysis · high effort56.1%Reading 66.3
- ProgramBenchVals AI · xhigh effortNear the floor0.5%Reading at most 65.7
- Code MigrationVals AI · xhigh effort45.2%Reading 64.4
- LiveBench CodingLiveBench · xhigh effort68.1%Reading 62.1
- Vibe Code BenchVals AI · xhigh effort69.8%Reading 57.2
- APEX-SWEMercor · xhigh effort37.0%Reading 47.9
- Terminal-Bench 4.0 (AA run)Artificial Analysis · high effort9.1%Reading 44.7
Research and reasoning
6 results · 64.6- FrontierMath Tier 4Epoch AI · xhigh effort72.5%Reading 74.5
- Mystery Game PuzzlesEpoch AI · high effort52.0%Reading 74.0
- LiveBench ReasoningLiveBench · xhigh effort92.8%Reading 74.0
- FrontierMath Tiers 1–3Epoch AI · xhigh effort85.3%Reading 73.4
- Humanity's Last Exam (AA run)Artificial Analysis · high effort45.0%Reading 63.3
- Chess PuzzlesEpoch AI · low effort26.0%Reading 40.7
Professional work
10 results · 57.9- LiveBench Data AnalysisLiveBench · xhigh effort81.6%Reading 80.5
- MedCodeVals AI · xhigh effort49.1%Reading 64.8
- APEX-AgentsMercor · xhigh effort55.1%Reading 63.0
- Legal Research BenchVals AI · xhigh effort40.4%Reading 62.4
- EMBVals AI · xhigh effort64.5%Reading 62.2
- τ-Bench Banking (AA run)Artificial Analysis · high effort36.7%Reading 58.6
- Finance AgentVals AI · xhigh effort51.8%Reading 57.1
- τ²-Bench Telecom (AA run)Artificial Analysis · high effort93.0%Reading 53.0
- Tax Agent BenchVals AI · xhigh effort21.8%Reading 50.1
- Harvey Legal Agent BenchmarkVals AI · xhigh effortNear the floor3.8%Reading at most 24.8
Knowledge and accuracy
6 results · 57.7- LiveBench LanguageLiveBench · xhigh effort87.4%Reading 82.8
- SimpleQA VerifiedEpoch AI · xhigh effort63.0%Reading 73.4
- LiveBench Instruction FollowingLiveBench · xhigh effort70.7%Reading 65.5
- AA-LCRArtificial Analysis · high effort84.3%Reading 57.1
- IFBench (AA run)Artificial Analysis · high effort71.6%Reading 44.8
- BullshitBenchBullshitBench · none effortCapped27.3%Reading 19.1
Human preference
1 result · 52.8- Arena TextLMArena · high effort1,481.4Reading 49.8
Shown, not scored
Composite indices, saturated or older evaluations, evaluations still being checked, and vision, writing and multilingual results. They do not move the score.
- LiveCodeBenchVals AI · CodingReference only85.3%
- MirrorCodeEpoch AI · CodingWatching10.0%
- SkillsBenchVals AI · CodingWatching62.2%
- SRE BenchVals AI · CodingWatching3.8%
- SWE-bench VerifiedVals AI · CodingReference only82.6%
- OTIS Mock AIMEEpoch AI · Research and reasoningReference only84.4%
- APEX-AccountingMercor · Professional workWatching6.2%
- CorpFinVals AI · Professional workReference only68.4%
- EBR-benchEpoch AI · Professional workWatching34.3%
- LegalBenchVals AI · Professional workReference only86.5%
- MedScribeVals AI · Professional workWatching86.9%
- TaxEvalVals AI · Professional workReference only75.0%
- GPQA DiamondEpoch AI · Knowledge and accuracyReference only90.7%
- GPQA DiamondVals AI · Knowledge and accuracyReference only93.2%
- MMLU-ProVals AI · Knowledge and accuracyReference only88.1%
- AA Intelligence IndexArtificial Analysis · Composite indicesReference only37.0
- LiveBench averageLiveBench · Composite indicesReference only80.2%
- Arena VisionLMArena · VisionReference only1,281.0
- MMMU ProVals AI · VisionReference only88.3%
No result yet
Scored evaluations this model has not taken. A missing result neither adds nor subtracts.
- Terminal-Bench 4.0 · Vals AI
- Vibe Code Bench 1–100 · Vals AI
- CyberBench Patch · Vals AI
- MysteryMechanism · Vals AI
- Terminal-Bench Science · Vals AI
- ProofBench · Vals AI