OpenAIReasoning modelGPT-5.1
Best configuration ranks #103 of 610 on the BenchLeader Index at 59.4 ±4.7. Last measured 8 Sept 2026. Released 12 Nov 2025.
- Blended price
- $3.44/M
- $1.25 in · $10.00 out
- Output speed
- 105 tok/s
- First answer
- 54 s
- first token 1.57 s
- Context
- 400k
- Overall index59
- Reasoning59
- Coding47
- Agents & tools56
- Knowledge67
- Instruction following71
- Human preference63
- Multimodal62
- Long context67
- Composite60
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Default” means the publisher did not say which setting was used.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| no reasoning | 46.6 | #366 | 101 tok/s | 1.17 s | $0.0035 | Agents & tools 53 · Composite 46 · Instruction following 45 · Knowledge 48 · Long context 48 · Maths 41 · Multimodal 47 · Reasoning 38 |
| low | – | – | 105 tok/s | 54 s | $0.0035 | Maths 52 · Reasoning 36 |
| medium | 52.8 | #239 | 105 tok/s | 54 s | $0.0035 | Agents & tools 54 · Coding 52 · Maths 62 · Reasoning 49 |
| high | 53.8 | #215 | 105 tok/s | 54 s | $0.0035 | Agents & tools 31 · Coding 55 · Human preference 65 · Knowledge 59 · Maths 56 · Multimodal 59 · Reasoning 55 |
| thinking | 55.2 | #181 | 105 tok/s | 54 s | $0.0035 | Instruction following 59 · Knowledge 61 · Multimodal 53 |
| defaultbest | 59.4 | #103 | 105 tok/s | 54 s | $0.0035 | Agents & tools 56 · Coding 47 · Composite 60 · Human preference 63 · Instruction following 71 · Knowledge 67 · Long context 67 · Multimodal 62 · Reasoning 59 |
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | no reasoning | low | medium | high | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|
| GPQA Diamond | 66.7%#164 | – | 85.0%#80 | 87.6%#59 | – | – | Epoch AI Benchmarking Hub | |
| Humanity's Last Exam | 6.8%#34 | – | – | – | – | 23.7%#15 | Scale AI / CAIS | |
| SimpleBench | – | – | – | 53.2%#41 | – | – | SimpleBench | |
| LMArena Hard Prompts | – | – | – | 1475#61 | – | 1456#88 | LMArena | |
| GPQA Diamond (AA)not in index | 64.3%#332 | – | – | – | – | 87.3%#99 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 5.3%#361 | – | – | – | – | 28.5%#115 | Artificial Analysis | |
| GPQA Diamond (Vals)not in index | – | – | – | 86.6%#39 | – | – | Vals AI | |
| ARC-AGI-1 | 5.8%#173 | 33.2%#142 | 57.7%#112 | 72.8%#95 | – | – | ARC Prize | |
| ARC-AGI-2 | 0.4%#171 | 1.9%#144 | 6.5%#113 | 17.6%#97 | – | – | ARC Prize |
Coding
| Benchmark | no reasoning | low | medium | high | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|
| SWE-bench Verified (Epoch) | – | – | – | 68.0%#26 | – | – | Epoch AI Benchmarking Hub | |
| SciCode | – | – | – | – | – | 43.3%#92 | SciCode | |
| WeirdML | – | – | – | 60.8%#44 | – | – | WeirdML | |
| GSO-Bench | – | – | – | 13.7%#16 | – | 13.7%#15 | GSO-Bench | |
| LMArena Coding | – | – | – | 1492#74 | – | 1475#97 | LMArena | |
| LMArena WebDev | – | – | 1392#74 | – | – | 1341#92 | LMArena | |
| LiveCodeBench | – | – | – | 86.5%#21 | – | – | Vals AI | |
| IOI | – | – | – | 21.5%#21 | – | – | Vals AI | |
| SWE-bench (Vals)not in index | – | – | – | 69.8%#61 | – | – | Vals AI | |
| SWE-bench Verified (bash only) | – | – | 66.0%#17 | – | – | – | SWE-bench | |
| SWE-bench Verified (any scaffold)not in index | – | – | 66.0%#21 | – | – | – | SWE-bench |
Agents & tools
| Benchmark | no reasoning | low | medium | high | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|
| Terminal-Bench | – | – | 47.6%#28 | – | – | 47.6%#28 | Terminal-Bench | |
| APEX-Agents | – | – | – | 17.5%#41 | – | 17.5%#41 | Mercor | |
| Terminal-Bench Hard | 22.7%#146 | – | – | – | – | 45.5%#32 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 46.5%#199 | – | – | – | – | 81.9%#111 | Artificial Analysis | |
| MCP Atlas | – | – | – | 50.1%#28 | – | – | Scale AI SEAL |
Maths
| Benchmark | no reasoning | low | medium | high | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|
| OTIS Mock AIME | 37.8%#182 | 63.9%#140 | 85.6%#81 | 88.6%#66 | – | – | Epoch AI Benchmarking Hub | |
| AIME (Vals) | – | – | – | 93.3%#15 | – | – | Vals AI | |
| MGSM | – | – | – | 93.0%#12 | – | – | Vals AI | |
| MathArena Apex | – | – | – | 1.0%#35 | – | – | MathArena |
Knowledge
| Benchmark | no reasoning | low | medium | high | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|
| SimpleQA Verified | – | – | – | 48.0%#27 | – | – | Epoch AI Benchmarking Hub | |
| AA-Omniscience | -34.5#232 | – | – | – | – | 5.4#70 | Artificial Analysis | |
| MMLU-Pro | – | – | – | 86.4%#41 | – | – | Vals AI | |
| LegalBench | – | – | – | 85.7%#17 | – | – | Vals AI | |
| CorpFin | – | – | – | 63.8%#40 | – | – | Vals AI | |
| TaxEval | – | – | – | 74.9%#25 | – | – | Vals AI | |
| MedQA | – | – | – | 96.4%#2 | – | – | Vals AI | |
| PRBench Finance | – | – | – | – | 48.0%#11 | – | Scale AI SEAL | |
| PRBench Legal | – | – | – | – | 49.3%#9 | – | Scale AI SEAL | |
| MultiNRC | – | – | – | – | 49.0%#10 | – | Scale AI SEAL |
Instruction following
| Benchmark | no reasoning | low | medium | high | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|
| IFBench | 43.2%#223 | – | – | – | – | 72.9%#45 | Artificial Analysis | |
| MultiChallenge | – | – | – | – | 63.4%#6 | – | Scale AI SEAL | |
| TutorBench | – | – | – | – | 54.1%#7 | – | Scale AI SEAL |
Human preference
| Benchmark | no reasoning | low | medium | high | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|
| LMArena Text | – | – | – | 1455#62 | – | 1439#80 | LMArena |
Multimodal
| Benchmark | no reasoning | low | medium | high | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|
| LMArena Vision | – | – | – | 1248#50 | – | 1235#59 | LMArena | |
| MMMU-Pro | 62.4%#162 | – | – | – | – | 75.5%#70 | Artificial Analysis | |
| VISTA | – | – | – | – | 43.8%#31 | – | Scale AI SEAL | |
| MMMU (validation) | – | – | – | – | – | 85.4%#1 | MMMU | |
| MMMU-Pro (official)not in index | – | – | – | – | 79.0%#4 | 76.0%#7 | MMMU |
Long context
| Benchmark | no reasoning | low | medium | high | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|
| AA-LCR | 45.0%#269 | – | – | – | – | 80.0%#51 | Artificial Analysis |
Composite
| Benchmark | no reasoning | low | medium | high | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|
| Epoch Capabilities Indexnot in index | – | – | – | – | – | 149.7#47 | Epoch AI Benchmarking Hub | |
| AA Intelligence Index | 13.3#246 | – | – | – | – | 24.7#120 | Artificial Analysis |
Where it wins
Benchmarks where this configuration ranks in the top five of every configuration measured.
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| OpenAI | 54 tok/s | 1.81 s | $1.25 | $10.00 | 400k | – |
| Azure | 41 tok/s | 1.26 s | $1.25 | $10.00 | 400k | – |
| OpenAI Flex | 24 tok/s | 1.57 s | $0.625 | $5.00 | 400k | – |
Price history
Listed price per 1M tokens over time, as recorded by OpenRouter for the provider with the longest history.
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $0.125 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0035 | $0.0032 | 56.9 s |
| Summarise a 30-page report | 12,000 / 600 | $0.021 | $0.011 | 59.7 s |
| Code edit | 6,000 / 1,500 | $0.022 | $0.017 | 1.1 min |
| Agentic coding session | 60,000 / 4,000 | $0.115 | $0.064 | 1.5 min |
| Structured extraction | 2,000 / 200 | $0.0045 | $0.0028 | 55.9 s |
See also
Data as of 9 Sept 2026. Compare these configurations.