MiniMax-M2 Instruct: pricing, context window and benchmarks
MiniMax-M2 Instruct is a multimodal model from MiniMax. It offers a 128,000-token context window, costs $5.11 per million input tokens and $25.04 per million output tokens, and averages 79.4 across the index's benchmark suite.
Pricing
MiniMax-M2 Instruct costs $5.11 per million input tokens and $25.04 per million output tokens. A blended 3:1 input-to-output mix works out to about $10.09 per million tokens. Cached input is listed at $0.72 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $5.11 | per 1M tokens |
| Output | $25.04 | per 1M tokens |
| Cached input | $0.72 | per 1M tokens |
| Blended 3:1 | $10.09 | estimate per 1M tokens |
| 1M in + 1M out | $30.15 | estimate |
Benchmarks
Scores below are modelled estimates on a 0–100 scale. Use them to compare relative position, not as official vendor results.
| Benchmark | Score | Rank | Percentile | Relative |
|---|---|---|---|---|
| MMLU-Pro | 82.1 | 121 | 80th | |
| GPQA Diamond | 73.7 | 178 | 70th | |
| SWE-bench Verified | 69.9 | 113 | 81th | |
| HumanEval | 86.7 | 95 | 84th | |
| MATH-500 | 82.1 | 165 | 73th | |
| AIME 2025 | 77.8 | 118 | 80th | |
| LiveCodeBench | 71.1 | 170 | 72th | |
| IFEval | 87.3 | 126 | 79th | |
| TruthfulQA | 80.0 | 178 | 70th | |
| HellaSwag | 88.5 | 125 | 79th | |
| MMMU | 74.1 | 189 | 69th | |
| Arena Elo | 1442.2 | 95 | 84th |
What MiniMax-M2 Instruct is good for
MiniMax-M2 Instruct is tuned toward coding work. It is a reasonable candidate for code generation, review, debugging and repository-scale tasks, especially when paired with a retrieval layer. At $5.11 per million input tokens, it sits in the mid-range of the index, and the 128,000-token context window is enough for large files and multi-file diffs.
Strengths
- Large 128,000-token context window for long documents and agent traces.
- Open weights make self-hosting, fine-tuning and data-residency review possible.
- Accepts more than text, including image input where the provider supports it.
- Coding benchmarks are a relative strength within its tier.
Trade-offs
- No standout weakness in the indexed specifications, but real-world behaviour still needs testing.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| Qwen3-Coder | Alibaba Cloud | 200000 | $4.94 | 79.5 |
| Claude Haiku 4.5 | Anthropic | 1000000 | $13.86 | 79.4 |
| Doubao 1.5 Pro Turbo | ByteDance | 128000 | $13.81 | 79.4 |
| MiniMax-M2 Instruct | MiniMax | 128000 | $5.11 | 79.4 |
| Nova Pro | Amazon | 32768 | $12.75 | 79.4 |
| Nova Pro Reasoning | Amazon | 200000 | $13.48 | 79.4 |
FAQ
How much does MiniMax-M2 Instruct cost?
MiniMax-M2 Instruct is listed at $5.11 per million input tokens and $25.04 per million output tokens. Cached input is $0.72 per million tokens. A blended 3:1 input-to-output mix is about $10.09 per million tokens.
What is the context window of MiniMax-M2 Instruct?
MiniMax-M2 Instruct has a context window of 128,000 tokens and can emit up to 4,096 output tokens per request in this index.
Is MiniMax-M2 Instruct open weights?
Yes. MiniMax-M2 Instruct is listed as an open-weight model under a open weights licence.
How does MiniMax-M2 Instruct score on benchmarks?
MiniMax-M2 Instruct averages 79.4 across the scored benchmarks in this index, ranking #137 of 600 models. Its strongest relative result is Arena Elo at 1442.2.
What is MiniMax-M2 Instruct best for?
MiniMax-M2 Instruct is tuned toward coding work. It is a reasonable candidate for code generation, review, debugging and repository-scale tasks, especially when paired with a retrieval layer. At $5.11 per million input tokens, it sits in the mid-range of the index, and the 128,000-token context window is enough for large files and multi-file diffs.