GPT-5.1

GPT-5.1 is a reasoning model from OpenAI. 28 benchmarks count toward its score, in 7 categories.

availableShows if the model has enough results for an index.
IndexOverall score. 50 is the middle.56.4 ±3.6
CoverageShare of the index weight with results.95%
SpeedOutput tokens per second.37/s
Input / 1MUS dollars per 1M input tokens.$1.25 batch $0.625
Output / 1MUS dollars per 1M output tokens.$10 batch $5 US dollars per 1M output tokens in a batch.
ContextMaximum tokens in one request.400K
EloLMArena rating and rank.1423 (#100)

50 is the middle of the board. The range shows the doubt in the index. Batch work costs less.

42,982 votes. Elo shows what people prefer. It does not change the score.

CapabilitiesScore per category. 50 is the middle.

50 is the middle
AgenticMulti-step tasks with tools.
52.2
CodingCode writing and repair.
54.5
ReasoningLogic problems and puzzles.
57.0
MultimodalTasks with images and text.
61.0
KnowledgeFacts and expert knowledge.
58.4
MultilingualTasks in many languages.
N/A
InstructionTasks with strict rules in the prompt.
63.1
MathMath problems.
58.7

Results

28 counted
BenchmarkThe test name.CategoryThe capability that the test measures.ResultThe score from the publisher.IndexThis result on the index scale.RunThe settings of the run.DateDate of the result.Published byThe source of the result.
AIMEMath93.3%57.216 Apr 2026Vals AI
MGSMMultilingual93.0%9 Jan 2026Vals AI
OTIS Mock AIME 2024-2025Math88.6%61.1high effortEpoch AI
GPQA diamondKnowledge87.6%59.1high effortEpoch AI
Artificial Analysis GPQA DiamondKnowledge87.3%58.0Artificial Analysis
GPQA DiamondKnowledge86.6%58.21 Sept 2026Vals AI
LiveCodeBenchCoding86.5%62.91 Sept 2026Vals AI
MMLU ProKnowledge86.4%56.61 Sept 2026Vals AI
MMMU ProMultimodal83.2%62.71 Sept 2026Vals AI
τ²-Bench Tool-Agent-User EvaluationAgentic81.9%57.6Victor Barres et al.
Artificial Analysis Long Context ReasoningReasoning80.0%63.6Artificial Analysis
Artificial Analysis MMMU-ProMultimodal75.5%59.4Artificial Analysis
Artificial Analysis IFBenchInstruction72.9%63.1Artificial Analysis
SWE-benchCoding69.8%53.51 Sept 2026Vals AI
SWE-Bench verifiedCoding66.9%51.2high effortEpoch AI
SWE-bench VerifiedCoding66.0%50.5medium effort · mini-SWE-agent1 Sept 2026SWE-bench team
Artificial Analysis Coding IndexCoding49.4%53.8Artificial Analysis
SimpleQA VerifiedKnowledge48.0%66.0high effortEpoch AI
Terminal-Bench 1.0Agentic47.5%51.812 Jan 2026Vals AI
Terminal-Bench 2.0Agentic44.9%54.84 Jun 2026Vals AI
Gert Labs Composite Game BenchmarkAgentic41.2%49.2Gert Labs
Artificial Analysis Omniscience AccuracyKnowledge37.7%60.5Artificial Analysis
Chess PuzzlesReasoning32.0%64.9high effortEpoch AI
FrontierMath-2025-02-28-PrivateMath31.0%59.6high effortEpoch AI
Artificial Analysis Humanity's Last ExamKnowledge28.5%55.5Artificial Analysis
Artificial Analysis Intelligence IndexKnowledge24.7%53.3Artificial Analysis
Vibe Code Bench v1.1Coding24.6%52.4OpenHands21 Sept 2026Vals AI
IOI v1Coding21.5%52.29 Aug 2026Vals AI
Mystery Game PuzzlesReasoning16.0%50.9medium effortEpoch AI
GDPval-AA normalizedAgentic15.6%47.5Artificial Analysis
FrontierMath-Tier-4-2025-07-01-PrivateMath12.5%57.0high effortEpoch AI
Critical Physics TasksReasoning4.9%48.7Artificial Analysis

28 benchmarks count, from 31 of 32 results. A grey row does not count. Too few models took that benchmark.

Sources

BenchLM benchmark aggregationUsed with attribution; per-benchmark results credited to their original authorsOpenRouter, collected directlyNo licence statedVals AI, collected directlyNo licence stated. Read from the public leaderboard and credited to Vals AIEpoch AI, collected directlyCC BY — free to use and redistribute with attributionSWE-bench team, collected directlyNo licence stated. The repository publishes submission records for reproducibility and transparency and asks that SWE-bench be cited

Same level, lower price

Qwen3.8-Flash-Next66.4 · Freedots3-note Preview65.5 · FreeApodex 1.1 Mini59.1 · Free

More from OpenAI

GPT-6 Astra83.5GPT-5.6 Sol76.7GPT-5.6 Terra73.4GPT-6 Sol76.9GPT-5.572.3