MiniMax M3
01
Snapshot
Overall
70
rank #13 of 22
Coding
68
rank #13 of 22
Multimodal
70
rank #18 of 22
Input / Output
$0.6 / $2.4
USD per 1M tokens
Context
1M
usable recall 90
Speed
~60 tok/s
TTFT 0.5s
Aggregated from public sources and independently weighted; methodology on the Terms page. Scores within 3 points are treated as statistically tied.
02
Specs & pricing
Every tracked field. List API prices in USD per 1M tokens; verify current vendor pricing before purchase.
| Vendor | MiniMax (CN) |
| Released / current | 2026.06 |
| Licensing | Open weights |
| Context window | 1M |
| Max output | 64K |
| Effective-context score | 90 |
| Input / output price | $0.6 / $2.4 per 1M tokens |
| Cache discount | No published cache discount |
| Free tier | Hailuo AI free; open-platform quota |
| Speed / TTFT | ~60 tok/s / 0.5s |
| Function calling | 84 |
| Refusal rate | ~7% |
| English / Chinese | 70 / 88 |
| Modalities | text |
| Fine-tuning | Yes |
| Private deployment | Yes (enterprise) |
| SOC2 / no-train | no / yes |
03
Strengths & watch-outs
Strengths
- Strong agent ability
- Open source
- Low price
Watch-outs
- Mid overall performance
- Text-only
04
Where it ranks among 22 models
Overall 70/100 (#13), coding 68 (#13), multimodal 70 (#18). See the full boards and side-by-side compare on the leaderboard.
05
Pricing in practice
Illustrative monthly bill for 100M input + 30M output tokens: $132 at list. Your mix and cache-hit ratio change this.
Illustrative model using public list rates before any enterprise agreement; cache discount: No published cache discount.
06
Frequently asked questions
How much does MiniMax M3 cost?
List API pricing is $0.6 input and $2.4 output per 1M tokens. No published cache discount. Hailuo AI free; open-platform quota
What is MiniMax M3's context window, and is it usable end to end?
It advertises a 1M window with 64K max output; its effective-context score is 90/100, which is the better predictor of whether long-document details are actually retained.
Is MiniMax M3 good for coding?
Its coding aggregate is 68/100, rank #13 of 22. It is adequate for scripts and assisted completion rather than the most demanding SWE-style work.
Can I self-host MiniMax M3?
Yes — it ships open weights and supports private deployment, so you can run it on your own infrastructure for data control; budget for GPU and operations.
How strong is it in Chinese and on multimodal inputs?
Chinese score is 88 versus English 70; multimodal aggregate is 70/100 (rank #18), accepting text.