MiniMaxMiniMax-M2.1
MiniMax-M2.1 is a MiniMax open-weights reasoning model, released 23 Dec 2025. Its best configuration ranks #265 of 777 on the BenchLeader Index at 53.9 ±3.1, in the upper half. The ± is the point: 271 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in instruction following (69) and lowest in reasoning (46). At $0.525 per million tokens blended it is mid-priced. Output speed of 63 tokens per second puts it slower than most, with a first answer in 33.3 s. Last measured 20 Oct 2026.
- Blended price
- $0.525/M
- $0.300 in · $1.20 out
- Output speed
- 63 tok/s
- measured by Artificial Analysis
- First answer
- 33 s
- first token 0.96 s
- Context
- 205k
- Full answer
- 41 s
- median, reasoning included
- Cost per run
- –
- Released
- 23 Dec 2025
- Overall index54
- Reasoning46
- Coding52
- Agents & tools50
- Maths54
- Knowledge52
- Instruction following69
- Human preference56
- Long context59
- Composite53
Versions
MiniMax has shipped 3 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| MiniMax M2.7 | 18 Mar 2026 | 55.7 | #219 |
| MiniMax M2.5 | 12 Feb 2026 | 53.7 | #276 |
| MiniMax-M2.1this page | 23 Dec 2025 | 53.9 | #265 |
How its index has moved
8 Sept 2026 to 18 Oct 2026The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. 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 | not stated | Source | Trend |
|---|---|---|---|
| Arena Hard Prompts | 1408#165 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 83.0%#157 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 23.2%#190 | Artificial Analysis | |
| GPQA Diamond (Vals)not in index | 78.5%#70 | Vals AI | |
| CritPt | 0.3%#264 | Artificial Analysis |
Coding
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Coding | 1437#158 | Arena (LMArena) | |
| Arena WebDev | 1384#98 | Arena (LMArena) | |
| LiveCodeBench | 81.8%#58 | Vals AI | |
| SWE-bench (Vals)not in index | 74.8%#45 | Vals AI | |
| ALE-Benchnot in index | 623.8#94 | Epoch AI Benchmarking Hub |
Agents & tools
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Terminal-Bench | 36.6%#51 | Terminal-Bench | |
| Terminal-Bench Hard | 28.8%#111 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 85.4%#86 | Artificial Analysis | |
| Terminal-Bench 2.0 (Vals)not in index | 37.1%#38 | Vals AI |
Maths
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AIME (Vals) | 77.9%#49 | Vals AI | |
| MGSM | 87.8%#53 | Vals AI | |
| MATH 500not in index | 89.0%#28 | Vals AI | |
| Arena Maths | 1391#161 | Arena (LMArena) |
Knowledge
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA-Omniscience | -32.9#285 | Artificial Analysis | |
| MMLU-Pro | 87.0%#35 | Vals AI | |
| LegalBench | 75.5%#110 | Vals AI | |
| CorpFin | 58.9%#79 | Vals AI | |
| TaxEval | 66.3%#108 | Vals AI | |
| MedQA | 91.2%#41 | Vals AI | |
| CaseLaw v2not in index | 55.8%#32 | Vals AI | |
| MedCodenot in index | 34.1%#79 | Vals AI | |
| MedScribenot in index | 80.8%#53 | Vals AI | |
| AA-Omniscience: accuracynot in index | 21.2%#280 | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 31.5%#213 | Artificial Analysis |
Instruction following
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| IFBench | 69.9%#67 | Artificial Analysis | |
| Arena Instruction Followingnot in index | 1385#154 | Arena (LMArena) |
Human preference
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Text | 1384#169 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1345#171 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1393#162 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1407#149 | Arena (LMArena) |
Long context
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA-LCR | 67.7%#234 | Artificial Analysis |
Composite
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA Intelligence Index v4.3.2 | 20.9#214 | Artificial Analysis |
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 |
|---|---|---|---|---|---|---|
| MiniMax | 50 tok/s | 0.56 s | $0.300 | $2.40 | 205k | fp8 |
| MiniMax | 45 tok/s | 0.96 s | $0.300 | $1.20 | 205k | fp8 |
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 $0.030 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0005 | $0.0004 | 38.1 s |
| Summarize a 30-page report | 12,000 / 600 | $0.0043 | $0.0019 | 42.8 s |
| Code edit | 6,000 / 1,500 | $0.0036 | $0.0024 | 57.1 s |
| Agentic coding session | 60,000 / 4,000 | $0.023 | $0.011 | 1.6 min |
| Structured extraction | 2,000 / 200 | $0.0008 | $0.0004 | 36.5 s |
See also
Data as of 21 Oct 2026. Compare with another model.
Cite as: BenchLeader, “MiniMax-M2.1: benchmarks, pricing, speed and rank”, https://www.benchleader.com/models/minimax-m2-1, data as of 21 Oct 2026.