xAIReasoning modelGrok 4.1
Best configuration ranks #169 of 610 on the BenchLeader Index at 55.9 ±4.0 (thinking reasoning effort). Last measured 8 Sept 2026. Released 27 Jun 2025.
- Blended price
- $1.56/M
- $1.25 in · $2.50 out
- Output speed
- –
- First answer
- –
- Context
- 2M
- Overall index56
- Reasoning59
- Coding46
- Agents & tools70
- Maths48
- Knowledge55
- Instruction following53
- Human preference66
- Multimodal50
- Long context63
- Composite55
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 | 47.9 | #342 | – | – | $0.0013 | Agents & tools 68 · Coding 31 · Knowledge 42 · Maths 40 · Reasoning 55 |
| thinkingbest | 55.9 | #169 | – | – | $0.0013 | Agents & tools 70 · Coding 46 · Composite 55 · Human preference 66 · Instruction following 53 · Knowledge 55 · Long context 63 · Maths 48 · Multimodal 50 · Reasoning 59 |
| default | 48.8 | #315 | – | – | $0.0013 | Agents & tools 46 · Coding 64 · Composite 43 · Human preference 65 · Instruction following 39 · Knowledge 40 · Long context 41 · Multimodal 32 · Reasoning 65 |
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 | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| SimpleBench | 56.0%#37 | 56.0%#37 | – | SimpleBench | |
| LMArena Hard Prompts | – | 1477#58 | 1474#63 | LMArena | |
| GPQA Diamond (AA)not in index | – | 85.3%#128 | 63.7%#334 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | – | 19.3%#165 | 5.1%#369 | Artificial Analysis | |
| GPQA Diamond (Vals)not in index | 65.2%#99 | 84.3%#51 | – | Vals AI |
Coding
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| LMArena Coding | – | 1499#64 | 1493#73 | LMArena | |
| LMArena WebDev | – | 1241#108 | – | LMArena | |
| LiveCodeBench | 42.6%#121 | 80.6%#64 | – | Vals AI | |
| IOI | 7.7%#36 | 3.1%#51 | – | Vals AI | |
| SWE-bench (Vals)not in index | – | 41.4%#81 | – | Vals AI |
Agents & tools
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| Cybench | – | – | 39.0%#6 | Cybench | |
| APEX-Agents | – | – | 12.8%#50 | Mercor | |
| Terminal-Bench Hard | – | 24.2%#137 | 14.4%#189 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | – | 93.3%#44 | 63.7%#167 | Artificial Analysis | |
| BFCL Overall | 58.3%#10 | 69.6%#5 | – | Berkeley Function Calling Leaderboard |
Maths
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| ProofBench | – | 4.0%#56 | – | Vals AI | |
| AIME (Vals) | 27.1%#73 | 91.9%#21 | – | Vals AI | |
| MGSM | 87.6%#57 | 89.5%#43 | – | Vals AI | |
| AIME 2026 | – | 94.2%#20 | – | MathArena | |
| HMMT February 2026 | – | 86.4%#20 | – | MathArena | |
| MathArena Apex | – | 5.2%#25 | – | MathArena |
Knowledge
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| AA-Omniscience | – | -29.9#215 | -50.9#325 | Artificial Analysis | |
| MMLU-Pro | 75.2%#108 | 84.2%#64 | – | Vals AI | |
| LegalBench | 79.1%#91 | 82.5%#63 | – | Vals AI | |
| CorpFin | 52.5%#97 | 66.0%#24 | – | Vals AI | |
| TaxEval | 61.9%#117 | 73.1%#51 | – | Vals AI | |
| MedQA | 76.0%#75 | 92.1%#34 | – | Vals AI |
Instruction following
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| IFBench | – | 52.7%#156 | 36.5%#285 | Artificial Analysis |
Human preference
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| LMArena Text | – | 1466#47 | 1459#52 | LMArena |
Multimodal
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| LMArena Vision | – | 1199#73 | – | LMArena | |
| MMMU-Pro | – | 63.3%#158 | 48.4%#210 | Artificial Analysis |
Long context
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| AA-LCR | – | 74.0%#120 | 31.3%#319 | Artificial Analysis |
Composite
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| AA Intelligence Index | – | 20.4#172 | 11.3#289 | Artificial Analysis |
Where it wins
Benchmarks where this configuration ranks in the top five of every configuration measured.
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 $0.050 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0013 | $0.0009 | – |
| Summarise a 30-page report | 12,000 / 600 | $0.017 | $0.0057 | – |
| Code edit | 6,000 / 1,500 | $0.011 | $0.0059 | – |
| Agentic coding session | 60,000 / 4,000 | $0.085 | $0.031 | – |
| Structured extraction | 2,000 / 200 | $0.0030 | $0.0012 | – |
See also
Data as of 9 Sept 2026. Compare these configurations.