MiniMax M2.5
MiniMax · Undisclosed (MoE)
Balanced across agents and multimodality
M2.5 continues MiniMax’s agent and long-context strengths with upper-mid vision. Built for multimodal workflow orchestration, not peak single-task recognition.
Last updated Oct 18, 2026Editorial scoring
Bottom line
Eight-dimension breakdown
editorial mapping · 0–100- MiniMax M2.583.8
Weights: perception 20 / documents 18 / charts 14 / video 14 / agents 14 / generation 6 / Chinese 6 / ecosystem 8. Definitions and sources on the methodology page.
Public benchmarks
as reported, not converted| Benchmark / metric | Value | Note |
|---|---|---|
| 上下文 | 512K | Official spec |
| 模态 | 4 | Text/image/video/audio |
Public scores come from different versions and prompts and are not interchangeable with our composite; they are one evidence input.
Scenario mapping
five stress tests, one scale| Scenario | Score | Observation |
|---|---|---|
| Chinese receipt OCR | 82 · Standard | |
| Chart data extraction | 82 · Standard | |
| 30-min video QA | 81 · Average | |
| Mobile GUI control | 86 · Strong tool use | |
| Chinese cultural context | 83 · General |
Strengths and limitations
Strengths
- Mature agent tool orchestration.
- 512K context is friendly to long material.
- Complete multimodal input coverage.
Limitations
- Vision and OCR are not standout.
- Few public benchmarks.
- Average enterprise compliance materials.
Price and real cost
List price
¥8/M tokens
≈ ¥8 / 1M output tokens. For 1,000 images at 1,024px, ~1,100 image tokens each and 300 output tokens, the monthly cost is about ¥4.8 (assuming input at 25% of output price).
Use your volumeAbout MiniMax M2.5
Is MiniMax M2.5 open source?
Licence: Closed commercial. No public weights; access is via vendor API or an on-prem licence.
How much does MiniMax M2.5 cost?
≈ ¥8 / 1M output tokens. It ranks #19 with a composite of 83.8. Real bills depend on resolution, tiling, caching and volume discounts.
What is MiniMax M2.5 best for?
Multimodal workflow orchestration; Content production agents; Long-material analysis。
Why does MiniMax M2.5 score 83.8?
The composite weights eight dimensions: perception 20%, documents 18%, charts 14%, video 14%, agents 14%, generation 6%, Chinese 6%, ecosystem 8%. Its strongest dimension is GUI agents & tool use; its weakest is Video & long-horizon reasoning.
Personal ranking · Not an official score
Every ranking, score and comparison on this page is an independent personal ranking, not an official score, and represents no vendor, institution or regulator. Scores are a subjective mapping of public evidence under our own weights — change the weights and the order changes. Disagree, or spotted a factual error? Contact the site owner: jizhi2024@foxmail.com — we check each report and log corrections in the changelog.