minimax-m3.1-flash API — per-token pricing and code examples
Price per 1M tokens
Cache rates apply when the provider reports cache tokens.
86% offAbout this model
minimax-m3.1-flash is served through OpenAI Chat Completions, OpenAI Responses, Anthropic Messages. You pay $0.021 per 1M input tokens and $0.084 per 1M output tokens. That is 86% below the official price of $0.15 and $0.60. Cache reads cost $0.0021 per 1M tokens when the provider reports cached tokens. It is included in budget packs from $1.99 up.
Where MiniMax M3.1 Flash fits
MiniMax calls M3.1 Flash "the latest M-series language model" and describes it as "an ultra-efficient open-weight model for everyday work." It is the generally available successor to the M3.1-Flash-Preview release. MiniMax designed minimax-m3.1-flash "for agentic reasoning, tool use, coding, and structured task execution," and also mentions assisted analysis for cybersecurity work. MiniMax gives M3.1 Flash a context window of 1,000,000 tokens. M3.1 Flash takes text, image and video input and produces text. MiniMax does not list a separate maximum output length, but it does say thinking tokens count toward max_tokens, so the limit you set has to cover both the reasoning and the answer.
MiniMax M3.1 Flash always thinks
M3.1 Flash reasons on every request, and MiniMax enforces that. Sending a thinking block with type disabled returns a 400 error, and so does an effort value of none. Depth is adjustable instead, across five levels: low, medium, high, xhigh and max. When you omit the setting, MiniMax uses max. In MiniMax's OpenAI-format reference the field is reasoning_effort; in its Anthropic-format reference the same choice is output_config.effort.
Two more limits catch people out on minimax-m3.1-flash. MiniMax's API does not support reasoning_split: false or service_tier. And because thinking shares the token budget, a small max_tokens can end the response with a max_tokens stop and no visible text at all. MiniMax also notes that in Claude Code, setting MAX_THINKING_TOKENS=0 has no effect on this model.
For tool calling, M3.1 Flash uses interleaved thinking. On simple tool calls, MiniMax says the model may call the tool directly without a thinking block, so check that the field exists before your code reads it. When reasoning content is present, keep it and send it back on the next turn.
What it costs on Cent AI
MiniMax M3.1 Flash is paid per token from your prepaid USD balance, at the input, cache read, cache write and output rates in the price box on this page. Since this model always thinks and defaults to max, reasoning tokens billed as output are part of every call. For high-volume, routine jobs, an explicit low or medium is the main lever on cost.
MiniMax lists M3.1 Flash among the models with automatic prompt caching, which applies once a request has 512 or more input tokens. Cent AI uses the cache rates only when MiniMax reports cached tokens in the usage of that request.
Using it from your tools
minimax-m3.1-flash is served on /v1/chat/completions, /v1/responses and /v1/messages. Cent AI does not change your parameters silently, so a disabled thinking block copied from another model is not removed and comes back as an upstream_error with status 400. The example below follows the Anthropic-format request in MiniMax's API reference:
{
"model": "minimax-m3.1-flash",
"max_tokens": 8192,
"output_config": { "effort": "medium" },
"messages": [{ "role": "user", "content": "Write unit tests for this parser." }]
}
In MiniMax's OpenAI-format reference the same setting is "reasoning_effort": "medium"; on /v1/chat/completions send that instead of the output_config block, and size max_completion_tokens with the same headroom for thinking.
Specification
- Context window
- —
- Input and output
- Text
- Endpoints
- /v1/chat/completions, /v1/responses, /v1/messages
- Smallest pack
- $1.99
Latency and error rate per model are on the status page. status.cent-ai.dev
What a request costs
Open the full calculatorCost per request
Estimate
minimax-m3.1-flashAn estimate at price version 1. Your bill comes from the usage the provider reports for each request.
Save $0.0294Code
curl https://api.cent-ai.dev/v1/chat/completions \
-H "Authorization: Bearer $CENT_AI_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "minimax-m3.1-flash", "messages": [{"role": "user", "content": "Hello"}]}'Set CENT_AI_KEY to a key from the dashboard. OpenAI SDKs use https://api.cent-ai.dev/v1; Anthropic SDKs use https://api.cent-ai.dev.
Price history
Version 1, in effect since Oct 30, 2026.
No published price change for this model yet.
Compare with
- deepseek-v4.1-flashInput$0.015
- gpt-6-lunaInput$0.01978
- qwen3.8-flashInput$0.015
- deepseek-v4-flashInput$0.015
Questions
How much does minimax-m3.1-flash cost on Cent AI?
$0.021 per 1M input tokens and $0.084 per 1M output tokens. Cache reads cost $0.0021 and cache writes $0.021 per 1M tokens. Each request is rounded up once.
Is there a minimum charge?
A successful request costs at least $0.00002. A request that ends early, because the client left or the output was cut off, pays only for its tokens.
How do I pay?
Top up a prepaid USD balance with QRIS or USDT/USDC, or buy a budget pack once. Nothing renews automatically.
What if the provider does not report usage?
The request is billed from an estimate and marked estimated on the Requests page. An estimate never contains cache tokens.
Which endpoint do I call?
/v1/chat/completions, /v1/responses, /v1/messages. The OpenAI base URL is https://api.cent-ai.dev/v1.
Try minimax-m3.1-flash with one key
One prepaid balance for every model. Pay with QRIS or USDT/USDC.