MiniMax M3
MiniMax · released May 31, 2026
MiniMax-M3 is a multimodal foundation model from MiniMax. It supports text, image, and video inputs with text output, a 1M-token context window, and is suited for long-horizon agentic work, coding,...
Specification
- Context window
- 1.05M
- Max output
- 512K
- Knowledge cutoff
- Not stated
- Parameters
- 427B
- Licence
- other
- Serving providers
- 13
- Moderated
- No
- Uptime
- 100.0%
Intelligence
29.2
82th percentile
Coding
58.6
Coding Index
Agentic
29.5
Agentic Index
Output speed
115 t/s
Median across providers
Latency
613ms
Time to first token
Cost per task
$0.11
Estimated
Benchmarks
Where the score comes from
The Intelligence Index is a composite. These are the underlying evaluations this model was actually measured on.
Evaluation scores
Percentage correct · higher is better
- GPQA Diamond92.9%
- τ²-bench (Telecom)88.9%
- AA-LCR (long context)83.0%
- IFBench82.9%
- SciCode47.1%
- Terminal-Bench Hard42.4%
- Humanity's Last Exam39.0%
An evaluation missing from this list was not run for this model — it is not a zero.
View as table
| Evaluation | Score |
|---|---|
| GPQA Diamond | 92.9% |
| τ²-bench (Telecom) | 88.9% |
| AA-LCR (long context) | 83.0% |
| IFBench | 82.9% |
| SciCode | 47.1% |
| Terminal-Bench Hard | 42.4% |
| Humanity's Last Exam | 39.0% |
Against its peers
Intelligence Index · this model highlighted, nearest peers in grey
- Ring-2.6-1TInclusionAI31.1
- GPT-5.2OpenAI30.4
- Gemini 3.1 Pro PreviewGoogle29.7
- Qwen3.7 MaxQwen29.5
- MiniMax M3MiniMax29.2
- GPT-5.2-CodexOpenAI28.5
- Solar Pro 4Upstage28.2
- GLM 5Z.ai27.9
Peers are the models sitting closest on the Intelligence Index — the set you would realistically choose between.
View as table
| Model | Intelligence |
|---|---|
| Ring-2.6-1T | 31.1 |
| GPT-5.2 | 30.4 |
| Gemini 3.1 Pro Preview | 29.7 |
| Qwen3.7 Max | 29.5 |
| MiniMax M3 | 29.2 |
| GPT-5.2-Codex | 28.5 |
| Solar Pro 4 | 28.2 |
| GLM 5 | 27.9 |
Percentile among all indexed models
Pricing
What it costs to run
List prices per million tokens, plus what one representative task works out to.
List price
- Input / 1M tokens
- $0.3
- Output / 1M tokens
- $1.2
- Cached input / 1M
- $0.06
- Blended 3:1
- $0.525
One task, estimated
$0.11
- Input tokens
- 50,000
- Output tokens
- 80,000
- Profile
- Reasoning
Estimated from list pricing: 50K input tokens plus 80K output tokens for reasoning models (25K for non-reasoning).
Arena
Head-to-head generation quality
Elo from pairwise judgements, broken out by the kind of thing the model was asked to build.
- Overall Elo
- 1,259
- Win rate
- 51.8%
- Strongest at
- Websites
- Tournaments
- 16,152
Elo by category
Dot position on a 1,150–1,275 scale · Elo has no meaningful zero
- 1,259
- 1,248
- 1,235
- 1,214
- 1,209
- 1,160
- 1,197
- 1,166
Agent categories (full-stack apps, mobile apps) are only scored for models tested in agent harnesses.
View as table
| Category | Elo | Win rate |
|---|---|---|
| Websites | 1259 | 51.8% |
| UI components | 1248 | 50.9% |
| Data visualisation | 1235 | 51.3% |
| Game development | 1214 | 45.9% |
| 3D scenes | 1209 | 50.2% |
| SVG | 1160 | 45.2% |
| Full-stack apps | 1197 | 47.3% |
| Mobile apps | 1166 | 45.9% |