MMiniMax · Released Feb 12, 2026
MiniMax M2.5
MiniMax-M2.5 is a SOTA large language model designed for real-world productivity. Trained in a diverse range of complex real-world digital working environments, M2.5 builds upon the coding expertise of M2.1...
- Released
- Feb 12 2026
- Context window
- 205K 33K out
- Input / 1M tokens
- $0.27 from Free
- Output / 1M tokens
- $0.95 from Free
Where to run it
52 providers compared
Free to try on OpenCode — free endpoints are often limited-time and rate-limited. Included in Alibaba Coding Plan (China), MiniMax Token Plan (minimax.io), Tencent Coding Plan (China), SCNet Token Plan, Alibaba Token Plan +3 more. Cheapest: DInference at $0.22 / $0.88 per 1M input / output tokens — 63% less than the priciest option.
| Provider | Input | Output | Cache read | Context | Max output | Uptime | vs. best |
|---|---|---|---|---|---|---|---|
| OpenCodefree · may be limited-time | Free | Free | Free | 205K | 131K | — | Free |
| Alibaba Coding Plan (China)subscription plan | Plan | Plan | Plan | 197K | 25K | — | Plan |
| MiniMax Token Plan (minimax.io)subscription plan | Plan | Plan | Plan | 205K | 131K | — | Plan |
| Tencent Coding Plan (China)subscription plan | Plan | Plan | Plan | 205K | 33K | — | Plan |
| SCNet Token Plansubscription plan | Plan | Plan | Plan | 205K | 131K | — | Plan |
| Alibaba Token Plansubscription plan | Plan | Plan | Plan | 197K | 33K | — | Plan |
| MiniMax Token Plan (minimax.cn)subscription plan | Plan | Plan | Plan | 205K | 131K | — | Plan |
| Alibaba Coding Plansubscription plan | Plan | Plan | Plan | 197K | 25K | — | Plan |
| Alibaba Token Plan (China)subscription plan | Plan | Plan | Plan | 197K | 33K | — | Plan |
| DInference | $0.22 | $0.88 | — | 200K | 32K | — | Cheapest |
| Deep Infra | $0.15 | $1.15 | $0.03 | 197K | 131K | — | +4% |
| Venice | $0.27 | $0.95 | $0.03 | 198K | 33K | 94.9%24h | +14% |
| Venice AI | $0.27 | $0.95 | $0.03 | 198K | 33K | — | +14% |
| StreamLake | $0.27 | $1.08 | $0.027 | 200K | 128K | 100%24h | +23% |
| NanoGPT | $0.27 | $1.08 | $0.027 | 205K | 131K | — | +23% |
| D.Run (China) | $0.29 | $1.16 | — | 205K | 131K | — | +32% |
| AtlasCloudFP8 | $0.29 | $1.20 | $0.06 | 197K | 177K | 100%24h | +35% |
| DigitalOcean | $0.30 | $1.20 | $0.06 | 66K | 59K | 100%24h | +36% |
| MinimaxFP8 | $0.30 | $1.20 | $0.03 | 205K | 131K | 100%24h | +36% |
| NovitaFP8 | $0.30 | $1.20 | $0.03 | 205K | 131K | 97.4%24h | +36% |
| GMICloudFP8 | $0.30 | $1.20 | $0.03 | 197K | 177K | 94.9%24h | +36% |
| FrogBot | $0.30 | $1.20 | $0.03 | 192K | 8K | — | +36% |
| Kilo Gateway | $0.30 | $1.20 | $0.03 | 198K | 33K | — | +36% |
| Vercel AI Gateway | $0.30 | $1.20 | $0.03 | 205K | 131K | — | +36% |
| LLM Gateway | $0.30 | $1.20 | $0.03 | 205K | 131K | — | +36% |
| HPC-AI | $0.30 | $1.20 | $0.03 | 196K | 195K | — | +36% |
| TokenGo | $0.30 | $1.20 | $0.03 | 205K | 131K | — | +36% |
| Baseten | $0.30 | $1.20 | — | 204K | 204K | — | +36% |
| Clarifai | $0.30 | $1.20 | — | 205K | 131K | — | +36% |
| Together AI | $0.30 | $1.20 | $0.06 | 205K | 131K | — | +36% |
| CloudFerro Sherlock | $0.30 | $1.20 | — | 196K | 16K | — | +36% |
| SiliconFlow | $0.30 | $1.20 | $0.03 | 197K | 131K | — | +36% |
| 302.AI | $0.30 | $1.20 | — | 205K | 131K | — | +36% |
| ZenMux | $0.30 | $1.20 | $0.03 | 205K | 131K | — | +36% |
| Novita AI | $0.30 | $1.20 | $0.03 | 205K | 131K | — | +36% |
| Amazon Bedrock | $0.30 | $1.20 | — | 197K | 98K | — | +36% |
| Meganova | $0.30 | $1.20 | $0.15 | 205K | 131K | — | +36% |
| DevPass (LLM Gateway) | $0.30 | $1.20 | $0.03 | 229K | 131K | — | +36% |
| Hugging Face | $0.30 | $1.20 | $0.03 | 205K | 131K | — | +36% |
| MiniMax (minimax.io) | $0.30 | $1.20 | $0.03 | 205K | 131K | — | +36% |
| CommandCode | $0.30 | $1.20 | $0.03 | 200K | 131K | — | +36% |
| Merge Gateway | $0.30 | $1.20 | $0.03 | 205K | 8K | — | +36% |
| Ofox | $0.30 | $1.20 | $0.03 | 205K | 131K | — | +36% |
| OrcaRouter | $0.30 | $1.20 | $0.03 | 205K | 131K | — | +36% |
| OpenCode | $0.30 | $1.20 | $0.06 | 205K | 131K | — | +36% |
| TensorX | $0.30 | $1.20 | $0.075 | 197K | 66K | — | +36% |
| Alibaba (China) | $0.30 | $1.20 | — | 205K | 131K | — | +36% |
| Eden AI | $0.30 | $1.20 | $0.03 | 205K | 131K | — | +36% |
| SiliconFlow (China) | $0.30 | $1.22 | — | 192K | 131K | — | +38% |
| Qiniu | $0.30 | $1.22 | — | 205K | 128K | — | +38% |
| MiniMax (minimax.cn) | $0.31 | $1.25 | $0.031 | 205K | 131K | — | +43% |
| Cortecs | $0.35 | $1.39 | $0.075 | 196K | 196K | — | +58% |
| GreenPT | $0.49 | $2.47 | $0.12 | 205K | 131K | — | +156% |
| Minimaxhighspeed · FP8 | $0.60 | $2.40 | $0.06 | 205K | 131K | 100%24h | +173% |
Value
Price vs. performance
MiniMax M2.5 Other MiniMax Other labs Value frontier · Terminal-Bench vs. input price per 1M tokens (log scale) · Drag to pan, pinch or Ctrl + scroll to zoom, double-click to zoom in.
Benchmarks
6 benchmark scores
| Benchmark | Score | Rank | Leader |
|---|---|---|---|
| Terminal-BenchSoftware engineering · Tasks completed with the programs in a real terminal (best agent) | 42.7% | #22of 40 | GPT-5.584.7% |
| Vending-Bench 2Agents & computer use · Run a simulated vending business for a year (final balance) | $-23 | #60of 61 | GPT-6 Astra$15,515 |
| WebDev ArenaDesign & visual · People vote on which model builds the better website (Elo) | 1384 | #68of 90 | Claude Opus 5.51820 |
| ARC-AGI-2Reasoning & knowledge · Few-shot abstract reasoning on grid puzzles | 4.9% | #55of 75 | GPT-6 Astra95.0% |
| ARC-AGI-1Reasoning & knowledge · Pattern induction from input–output examples | 63.7% | #49of 75 | Gemini 4 Argon98.5% |
| ProofBenchMaths · Graduate maths proofs verified in Lean 4 | 4.0% | #69of 70 | Claude Sonnet 5.5100.0% |
Specs
Details
- Model ID
minimax/minimax-m2.5- Developer
- MiniMax
- Release date
- Feb 12, 2026 ·
- Context window
- 204,800 tokens
- Max output
- 32,768 tokens
- API price
- $0.27 / $0.95 per 1M input / output tokens
- Providers
- 52
- Input types
- text
- Reasoning
- Yes
- Tool calling
- Yes
- Open weights
- Yes
- Free tier
- No
Release date and prices from OpenRouter's catalog; providers from OpenRouter and models.dev. Buying directly from a provider can cost a little differently. Speed, latency and uptime are measured on OpenRouter only; uptime is the share of OpenRouter requests each provider served over the last day. Benchmark scores are the best result per benchmark from Epoch AI, licensed CC BY 4.0. Rank counts every model tracked here with a score.