MiniMax M2.5
Tool useReasoningPrompt cachingStructured output
Context
1M
Max output
197K
Serving providers
0
Cheapest input
n/a /1M
Benchmarks
independent evaluations · source: ARC Prize, Artificial Analysis and Hugging Face leaderboardsSpeed & latency
Output speed
74 t/s
tokens / second
Time to first token
5.18s
latency
Headline indices
Individual benchmarks · click a row for effort variants
| Benchmark | Best score | Effort | Cost / task | |
|---|---|---|---|---|
| ARC-AGI-2 | 5% | default | $0.170 | |
| CritPt | 1% | default | — | |
| EvalComputeProxy | 20.1 | default | — | |
| GPQA Diamond | 85% | default | — | |
| Humanity's Last Exam | 19% | default | — | |
| IFBench | 72% | default | — | |
| Long-Context Reasoning | 66% | default | — | |
| Omniscience | -39.7 | default | — | |
| SciCode | 43% | default | — | |
| SWE-bench Pro | 55% | default | — | |
| SWE-bench Verified | 76% | default | — | |
| TerminalBench Hard | 35% | default | — | |
| τ²-bench | 95% | default | — |
Pricing by serving provider
availabilityNo public price yet. Not offered on-demand at any tracked provider.
Price history
input + output $/1M since we started tracking Input Output
Frequently asked questions
MiniMax M2.5 pricing, context and availabilityWhat is the context window of MiniMax M2.5?
MiniMax M2.5 accepts up to 1M tokens of context and can return up to 197K output tokens.
What can MiniMax M2.5 do?
MiniMax M2.5 supports tool use, reasoning, prompt caching and structured output.
Other MiniMax models
compare pricing across the MiniMax lineup Pricing verified against LiteLLM + OpenRouter + provider pages · refreshed daily
Every weekday
AI moves fast. Here's your debrief.
News, analysis, tools, and more.
For people who build with AI