Skip to navigation
Authors and ModelsMiniMax

MiniMax M3.1 Flash

MiniMax M3.1 Flash
MiniMax · Text Generation
POST /v1/chat/completions

Fast multimodal coding and agent model with 1M context, adjustable reasoning depth, image and video input, function calling, and built-in web search.

At a glance

FieldValue
Model idminimax-m3-1-flash
Model release date2026-10-19
Input modalitiesText, Image, Video
Output modalitiesText
Context window1M
Weight precision-
Max output tokens524,288
RegionSingapore
Featuresreasoning, vision, video, web_search, function_calling, cache, long_context, agentic_coding
Native inferenceNo
NewYes
Structured outputNot supported
Supported endpointsPOST /v1/chat/completions, POST /v1/responses, POST /v1/messages, POST /v1beta/models/minimax-m3-1-flash:generateContent
Alternate model idsminimax-m3.1-flash, minimax/m3.1-flash, minimax/minimax-m3.1-flash

Pricing

ChargeSpecRate
Inputper 1M prompt tokens$0.1125 (was $0.15)
Outputper 1M generated tokens$0.45 (was $0.60)
Implicit cache readper 1M cached input tokens$0.01125 (was $0.015)
Web searchper call when invoked$0.01

Example request

curl https://api.empiriolabs.ai/v1/chat/completions \
  -H 'Authorization: Bearer $EMPIRIOLABS_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{"model": "minimax-m3-1-flash", "messages": [{"role":"user","content":"Hello"}]}'

Parameters

ParameterTypeRequiredDefaultDescription
reasoning_effortenumno"max"Reasoning is always on; this sets how much effort the model spends before answering. Higher effort uses more reasoning tokens. · Allowed: low, medium, high, xhigh, max
temperaturenumberno1Controls randomness. Lower values are more deterministic; higher values are more exploratory. · Range: 0 – 2
top_pnumberno0.95Controls nucleus sampling by limiting generation to the most likely token mass. · Range: 0 – 1
max_completion_tokensintegerno32768Maximum generated tokens, including reasoning tokens. · Range: 1 – 524288
stoparrayno-Optional stop sequence or list of stop sequences.
tool_web_searchbooleannofalseSearch the web for real-time information. Adds $0.01 to the request cost for each invoked call.
toolsarrayno-OpenAI-compatible tool definitions for function calling.
tool_choiceobjectno-Optional OpenAI-compatible tool_choice value.
disable_formattingbooleannofalseReturn raw provider-style output without EmpirioLabs source formatting where supported.

Notes

Text, image, and video input with text output, a 1M token context, and up to 524,288 output tokens. The model always reasons: reasoning_effort sets how deeply (low, medium, high, xhigh, or max; the default is max), and the reasoning text is returned in reasoning_content. Reasoning tokens are billed as output tokens. Repeated prompt prefixes are cached automatically and billed at the cached input rate. Set tool_web_search to true to let the model search the web; each executed search is billed, and one request can run several searches. Answers that used web search end with a Sources list unless disable_formatting is true. Structured output through response_format is not supported; ask for JSON in the prompt instead.

Per-tool billing (usage.tool_usage)

When this model invokes tools (web search, code interpreter, etc.) inside a single request, the response carries a normalized usage.tool_usage map alongside the token counts. The example below shows the shape, exact field names, units, and which tools appear can vary slightly per provider:

"usage": {
  "prompt_tokens": 123,
  "completion_tokens": 456,
  "cost_usd": 0.0042,
  "tool_usage": {"web_search": 3, "code_interpreter": 1}
}

The tool counts are already factored into cost_usd, they are surfaced for transparency so you can audit per-tool billing. The field is omitted when no tools were invoked.

Machine-readable schema: GET https://api.empiriolabs.ai/v1/models/minimax-m3-1-flash.