All models
MiniMax M3
minimax/minimax-m3
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,...
Modalities
Text
Image
Video
In / Out price
$0.30 / $1.20 per 1M
Context
1.05M · 512K out
Released
Jun 1, 2026
Providers
Live list prices and uptime across 8 providers serving this model.
| Provider | Input /M | Output /M | Cache read /M | Context | Max output | Uptime 24h |
|---|---|---|---|---|---|---|
| GMICloudfp8 | $0.24 | $0.96 | $0.048 | 1.05M | — | 98.21% |
| Novitafp8 | $0.30 | $1.20 | $0.06 | 1M | 131K | 98.72% |
| Venicefp8 | $0.30 | $1.20 | $0.06 | 524K | 66K | 95.96% |
| Minimaxfp8 | $0.30 | $1.20 | $0.06 | 524K | 512K | 99.04% |
| AtlasCloudfp8 | $0.30 | $1.20 | $0.06 | 524K | 524K | 98.39% |
| Together | $0.30 | $1.20 | $0.06 | 524K | — | 96.99% |
| Morph | $0.30 | $1.20 | — | 256K | 256K | 93.79% |
| DeepInfrafp8 | $0.30 | $1.20 | $0.06 | 524K | 512K | 97.37% |
No implicit prompt caching; cache reads require explicit cache control.
Capabilities
Tool calling
Reasoning
frequency_penalty
include_reasoning
logit_bias
logprobs
max_tokens
min_p
presence_penalty
reasoning
repetition_penalty
response_format
seed
stop
structured_outputs
temperature
tool_choice
top_k
top_logprobs
top_p
Pricing detail
- Input
- $0.30
- Output
- $1.20
- Cache read
- $0.06
- Cache write
- —
USD per 1M tokens.