Submit
MiniMaxWeights not listed·1000K context·— params

MiniMax-M3.1-Flash-Previewminimax/minimax-m3-1-flash-preview

MiniMax's multimodal coding preview supports a 1M context and adjustable reasoning effort. Current official access is through M Plan and MiniMax Code. No pay-as-you-go rate, downloadable checkpoint or maximum output limit has been verified for this exact preview.

Official sources checked October 2, 2026. Evaluation suggestions below are not benchmark results.

Subscription access is not a per-token offer

The official model catalog limits this preview to M Plan and MiniMax Code for now. The September 28 Code changelog records its availability; an exact first-launch date has not been independently verified here. Use the exact model ID MiniMax-M3.1-Flash-Preview with the credential type required by your plan.

There is no verified pay-as-you-go row on this page. Do not substitute M3's token price or divide a monthly subscription by an assumed token allowance. Confirm access and current quota in your account. Downloadable weights and a license for this preview have not been verified.

Reasoning effort changes the migration

The Anthropic-compatible API documents always-on thinking: disabled thinking and effort none return an error. output_config.effort accepts low, medium, high, xhigh and max, defaulting to max. Reasoning counts toward max_tokens, so an output cap can be reached before a final text answer appears.

Treat an empty final answer differently from a successful short answer. Log stop reasons, reasoning usage and final text separately. Preserve thinking blocks unchanged across tool turns as the API documentation instructs; an adapter that discards them needs review before migration.

Evaluate effort on the tasks you actually run

Use a bounded code change, a multi-file defect and a screenshot-assisted repair as separate cases. Run each at two effort levels with the same tools and acceptance criteria. Compare tests passed, unrelated changes, retries, time to an acceptable result and plan usage. These are suggested checks, not tests completed by this directory.

Higher effort may be unnecessary for a routine transformation but useful for an ambiguous defect. Decide from the observed task outcomes. A context ceiling alone does not prove reliable retrieval from a repository; include a case whose correct fix requires evidence from distant files.

Official source 1 →Official source 2 →Official source 3 →Official source 4 →

Pricing across providers

No current verified API pricing is listed. See the official model documentation for availability and billing.

Capabilities

  • reasoning
  • coding
  • tool_use
  • multimodal_input

Technical specs

Context
1000K
Max output
—
Parameters
—
Release
—
Training cutoff
—
License
—
Official documentation →

Similar models

Compare with

  • MiniMax-M3.1-Flash-Preview vs DeepSeek V4.1 Flash
    Comparison planned — not yet published
  • MiniMax-M3.1-Flash-Preview vs GLM 5.3 Flash
    Comparison planned — not yet published
  • MiniMax-M3.1-Flash-Preview vs Kimi K2.5
    Comparison planned — not yet published

Frequently asked

How much does MiniMax-M3.1-Flash-Preview cost?+
No current verified API pricing is listed. This does not mean the model is free or that a hosted API is currently available. Check the official model documentation below.
How do I access MiniMax-M3.1-Flash-Preview from outside China?+
Availability, account verification and payment methods vary by provider and region. Check the official documentation and the selected provider's access requirements before integrating. A directory listing does not guarantee global API access.
Is MiniMax-M3.1-Flash-Preview open-source?+
Downloadable weights are not listed in this directory for MiniMax-M3.1-Flash-Preview. See the official documentation and availability notes for API, subscription or planned weight-release details.
What's the maximum context window for MiniMax-M3.1-Flash-Preview?+
The model supports up to 1,000,000 tokens of context (input + output). Some hosted versions may impose a smaller limit — check the "Context" column in the pricing matrix for each provider.