MiniMax › MiniMax M2.5 › Pricing
$0.30 / $1.20per 1M tokens in / out
$0.03cache read · −90%
#190on OpenRouter by tokens · 11.6B / week
Price sheetper 1M tokens
| Input | $0.30 | list |
| Output | $1.20 | |
| Cache read | $0.03 | −90% vs input |
| Context · max output | 205K · 131K | tokens |
Where to buy10 providers · per 1M tokens
| Provider | Input | Output | Context |
|---|---|---|---|
| Azure AI Foundry | $0.033 | $1.32 | — |
| DeepInfra | $0.15 | $1.15 | — |
| AtlasCloud | $0.29 | $1.20 | 197K |
| AWS Bedrock | $0.30 | $1.20 | — |
| DigitalOcean | $0.30 | $1.20 | 66K |
| Friendli | $0.30 | $1.20 | 197K |
| GMICloud | $0.30 | $1.20 | 197K |
| Minimaxlab | $0.30 | $1.20 | 205K |
| Novita | $0.30 | $1.20 | 205K |
| lab-qwen | $0.30 | $1.21 | — |
In MiniMax's line-upper 1M in / out
| MiniMax M3.1 Flash | $0.15 / $0.60 | −50% |
| MiniMax 01 | $0.20 / $1.10 | −19% |
| MiniMax M2-her | $0.30 / $1.20 | 0% |
| MiniMax M2.5 | $0.30 / $1.20 | this model |
| MiniMax M3 | $0.30 / $1.20 | 0% |
| MiniMax M1 | $0.40 / $2.20 | +62% |
| MiniMax M2.7-highspeed | $0.60 / $2.40 | 2× |
Cheaper for the same workvs MiniMax M2.5
| Claude Haiku 5.5 max | -62% | score +90% | $0.21 |
| MiniMax M3.1 Flash thinking | -50% | score +87% | $0.22 |
| GLM 5.3 Flash maxOpen | -55% | score +83% | $0.28 |
| Ling 3.1 FlashOpen | -14% | score +80% | $1.00 |
| Qwen3.8 Flash thinkingOpen | -56% | score +75% | $0.37 |
Pricing questions
How much does MiniMax M2.5 cost?›
$0.30 per 1M input tokens and $1.20 per 1M output tokens. Cached input is $0.03.
Is MiniMax M2.5 free to use?›
There is no free API tier, but the weights are open: you can run it on your own hardware for the cost of compute.
How much does prompt caching save?›
Cache reads cost $0.03 per 1M, 90% below the input price.
Is MiniMax M2.5 cheaper than MiniMax M2.1?›
It costs the same per token: $0.30 / $1.20.
What is a cheaper alternative to MiniMax M2.5?›
Claude Haiku 5.5 at max effort costs -62% per task with a score +90% on Humanity's Last Exam.