MiniMax M2.7
7 cheaper ranked models score higher: DeepSeek V4.1 Flash (59.0, $0.518), GPT-5.6 Luna (58.5, $0.450), Inkling 2 Small (58.4, $0.475) and 4 more.
As of 8 Nov 2026, MiniMax M2.7 by MiniMax scores 56.6 on the BenchLeader Index (593 ranked configurations), with its best results in instruction following (74) and weakest in agents & tools (47). API pricing: $0.300 per 1M input tokens and $1.20 per 1M output tokens ($0.525 blended at 3:1), list price on 8 Nov 2026. It streams about 49 tokens/s (Artificial Analysis). Released 18 Mar 2026.
Key facts (as of 8 Nov 2026)
- MiniMax M2.7 API pricing: $0.300 per 1M input tokens and $1.20 per 1M output tokens ($0.525 blended at 3:1) (list price, 8 Nov 2026).
- MiniMax M2.7 streams about 49 tokens/s (Artificial Analysis, 8 Nov 2026).
- MiniMax M2.7 was released on 18 Mar 2026 (MiniMax).
In detail. MiniMax M2.7 is a MiniMax open-weights reasoning model, released 18 Mar 2026. Its best configuration ranks #173 of 593 on the BenchLeader Index at 56.6 ±4.8, in the upper half. The ± is the point: 264 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 (74) and lowest in agents & tools (47). At $0.525 per million tokens blended it is cheaper than most ranked models. Output speed of 49 tokens per second puts it in the slowest quarter, with a first answer in 24.3 s. Last measured 6 Nov 2026.
- Blended price
- $0.525/M
- $0.300 in · $1.20 out
- via Artificial Analysis
- Output speed
- 49 tok/s
- measured by Artificial Analysis
- First answer
- 24 s
- first token 0.85 s
- Context
- 205k
- via models.dev
- Full answer
- 88 s
- median, reasoning included
- Cost per task
- –
- Tokens per task
- –
- Released
- 18 Mar 2026
- Overall index57
- Reasoning60
- Coding51
- Agents & tools47
- Maths54
- Knowledge52
- Instruction following74
- Human preference59
- Long context64
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 M2.7this page | 18 Mar 2026 | 56.6 | #173 |
| MiniMax M2.5 | 12 Feb 2026 | 53.7 | #244 |
| MiniMax-M2.1 | 23 Dec 2025 | 55.0 | #208 |
How its index has moved
8 Sep 2026 to 8 Nov 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 | 1441#128 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 87.4%#100 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 29.6%#162 | Artificial Analysis | |
| GPQA Diamond (Vals)not in index | 86.6%#40 | Vals AI | |
| CritPtnot in index | 0.6%#266 | Artificial Analysis |
Coding
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| SciCode | 50.1%#117 | SciCode | |
| WeirdML | 37.0%#121 | WeirdML | |
| Arena Coding | 1481#108 | Arena (LMArena) | |
| Arena WebDev | 1398#100 | Arena (LMArena) | |
| SciCode (AA)not in index | 50.1%#116 | Artificial Analysis | |
| LiveCodeBench | 79.9%#71 | Vals AI | |
| SWE-bench (Vals)not in index | 73.8%#50 | Vals AI | |
| Code Migrationnot in index | 8.7%#65 | Vals AI | |
| ProgramBenchnot in index | 0.0%#33 | Vals AI | |
| Vibe Code Bench v1.1not in index | 11.9%#103 | Vals AI | |
| ALE-Benchnot in index | 599.3#102 | Epoch AI Benchmarking Hub | |
| Terminal-Bench 4.0 (AA)not in index | 0.0%#171 | Artificial Analysis | |
| Terminal-Bench 2.1 (AA)not in index | 55.4%#109 | Artificial Analysis | |
| Design Arena Websitenot in index | 1240#66 | Design Arena |
Agents & tools
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Terminal-Bench | 45.1%#40 | Terminal-Bench | |
| Terminal-Bench Hard | 39.4%#57 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 84.8%#98 | Artificial Analysis | |
| Terminal-Bench 2.1 (Vals) | 48.7%#63 | Vals AI | |
| τ³-Banking (AA) | 9.9%#147 | Artificial Analysis | |
| ITBench SRE (AA)not in index | 26.5%#36 | Artificial Analysis | |
| Analyst Agent (AA)not in index | 11.3%#46 | Artificial Analysis | |
| APEX-Agents (AA)not in index | 10.6%#29 | Artificial Analysis | |
| Finance Agent v2not in index | 27.9%#75 | Vals AI | |
| Harvey's Legal Agent Benchmark v1not in index | 0.0%#64 | Vals AI | |
| Legal Research Benchnot in index | 10.6%#78 | Vals AI | |
| Tax Agent Benchnot in index | 6.9%#73 | Vals AI | |
| Terminal-Bench 2.0 (Vals)not in index | 47.2%#22 | Vals AI | |
| ExploitBenchnot in index | 13.3%#9 | Epoch AI Benchmarking Hub | |
| GBAEvalnot in index | 0.0%#22 | Epoch AI Benchmarking Hub | |
| AutomationBenchnot in index | 4.4%#188 | Artificial Analysis | |
| GDP.pdf (AA)not in index | 6.0%#172 | Artificial Analysis |
Maths
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| ProofBench | 3.0%#79 | Vals AI | |
| AIME (Vals) | 91.0%#26 | Vals AI | |
| Arena Maths | 1423#122 | Arena (LMArena) |
Knowledge
Instruction following
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| IFBench | 75.7%#27 | Artificial Analysis | |
| Arena Instruction Followingnot in index | 1408#131 | Arena (LMArena) |
Human preference
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Text | 1415#136 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1364#158 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1426#122 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1432#122 | Arena (LMArena) |
Long context
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA-LCR v1.1 | 78.3%#120 | Artificial Analysis | |
| MLCRnot in index | 5.6%#144 | Artificial Analysis |
Composite
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA Intelligence Index v4.3.2not in index | 22.8#205 | Artificial Analysis | |
| AA Progress Index v1not in index | 272.6#195 | Artificial Analysis |
API pricing: 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 |
|---|---|---|---|---|---|---|
| Groq | 292 tok/s | 0.25 s | $0.600 | $1.80 | 197k | – |
| GMICloud | 45 tok/s | 1.29 s | $0.210 | $0.840 | 197k | fp8 |
| AtlasCloud | 45 tok/s | 1.31 s | $0.300 | $1.20 | 197k | fp8 |
| NovitaAI | 36 tok/s | 1.43 s | $0.270 | $1.08 | 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 $0.060 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0005 | $0.0004 | 30.5 s |
| Summarize a 30-page report | 12,000 / 600 | $0.0043 | $0.0022 | 36.6 s |
| Code edit | 6,000 / 1,500 | $0.0036 | $0.0025 | 55.2 s |
| Agentic coding session | 60,000 / 4,000 | $0.023 | $0.012 | 1.8 min |
| Structured extraction | 2,000 / 200 | $0.0008 | $0.0005 | 28.4 s |
See also
Embed or cite
Show where MiniMax M2.7 ranks on a model card or README. The badge redraws itself with every daily update, carries the data date and links back to this page. To see ranks like this on OpenRouter and Hugging Face model pages, get the Chrome extension.
[](https://www.benchleader.com/models/minimax-m2-7)Cite this
The ranking changes daily, so a reference carries the data date. APA style:
BenchLeader. (2026, November 8). MiniMax M2.7 on the BenchLeader Index (method v1.4) [Data set]. Retrieved November 8, 2026, from https://www.benchleader.com/models/minimax-m2-7
BibTeX:
@misc{benchleader_minimax_m2_7_20261108,
author = {{BenchLeader}},
title = {{MiniMax M2.7 on the BenchLeader Index (method v1.4)}},
year = {2026},
month = nov,
howpublished = {\url{https://www.benchleader.com/models/minimax-m2-7}},
note = {Data as of 2026-11-08}
}Every badge style, the embeddable top-10 widget and how to cite the Index
Questions people ask
- What are MiniMax M2.7's benchmark scores?
- As of 8 Nov 2026, MiniMax M2.7 scores 56.6 ±4.8 on the BenchLeader Index, strongest in instruction following (74) and weakest in agents & tools (47). That rests on independent benchmark results only, and every score on this page links to its publisher.
- How does MiniMax M2.7 rank?
- MiniMax M2.7 is #173 of 593 ranked configurations on 8 Nov 2026. 264 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 69.4.
- How much does MiniMax M2.7 cost per 1M tokens?
- As of 8 Nov 2026, MiniMax M2.7 costs $0.300 per 1M input tokens and $1.20 per 1M output tokens, $0.525 blended (three input tokens per output token).
- When was MiniMax M2.7 released?
- MiniMax released MiniMax M2.7 on 18 Mar 2026.
Data as of 8 Nov 2026. Compare with another model.
Cite as: BenchLeader, “MiniMax M2.7: benchmarks, pricing, speed and rank”, data as of 8 Nov 2026, https://www.benchleader.com/models/minimax-m2-7.