Model detail
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, and tool use. It is built on MiniMax Sparse Attention (MSA), which replaces full attention with KV-block selection to cut per-token compute at long context — roughly 1/20 the cost of the previous generation at 1M tokens, with substantially faster prefill and decode while retaining quality across most tasks. Trained as a native multimodal model on interleaved data and tuned for multi-turn, production-like collaboration via an interactive user-simulator framework, the model is oriented toward sustained, multi-step tasks rather than single-turn execution.
Model specs
- Model ID
-
MiniMax-M3
- Context length
- 1M
- Max output
- 262.144K
- I/O modalities
- Released
- 2026-05
02
API endpoints
One MixRoute gateway, OpenAI-compatible
-
OpenAI-compatible
/v1/chat/completionsPOST
03
Tiered pricing
Unit: /1M Tokens
| Tier | Input /1M Tokens | Output /1M Tokens | Cache read /1M Tokens |
|---|---|---|---|
| standard Length ≤ 512K | $0.6000 | $2.4000 | $0.1200 |
| long_context Length > 512K | $1.2000 | $4.8000 | $0.2400 |
05
Selection summary
Quickly judge whether it fits your workload
What is this model good for?
Multimodal foundation model from MiniMax
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, and tool use.
What should you check before using it?
1,000,000 context, 262,144 output tokens
MiniMax model that supports reasoning. Knowledge cutoff: 2026-01. Released 2026-06-01.
Why use it through MixRoute?
MiniMax-M3 runs tools and structured output here
Tool calling and structured output run on that route as well. MixRoute serves MiniMax-M3 on the OpenAI-compatible endpoint /v1/chat/completions; point your existing client at https://api.mixroute.ai/v1 with the model ID MiniMax-M3 and re-run your integration tests.
07
Capabilities & compatibility
Unconfirmed capabilities stay "Unknown," no guesses
Tool calling
MiniMax-M3 supports function and tool calls.
tools definition; the response returns tool_calls to execute and feed results back.
Structured output
MiniMax-M3 can produce JSON-formatted output on demand.
response_format: { type: "json_schema" } forces valid JSON output.
Streaming
MiniMax-M3 supports streamed responses.
stream: true streams tokens incrementally, so chat UIs render as they arrive.
Prompt caching
MiniMax-M3 supports prompt caching under supported routing conditions.
Vision input
MiniMax-M3 can process image inputs alongside text.
image_url content part alongside text messages.
OpenAI SDK
Call MiniMax-M3 through MixRoute's OpenAI-compatible endpoint.
base_url to https://api.mixroute.ai/v1; everything else follows the OpenAI style.
08
Token cost estimator
Live estimate from this page's pricing, not an actual bill
Share of the same prefix read repeatedly, up to 100%
09
FAQ
Reasoning tokens are billed as output on MiniMax-M3: $2.4000 per 1M tokens for inputs up to 512,000 tokens, and $4.8000 per 1M tokens above that. They are generated before the visible answer and count toward both latency and spend, so a higher reasoning setting costs more even when the final answer is short.
One endpoint, a testable decision
Test this model and alternate routes with the same request format
Start from a real workload, then let quality, total cost, and failure conditions decide whether to send production traffic.