Discussion

MiniMax M3.1 Flash offers eight times Gemini 3.6 Flash’s output ceiling

In Model Chat

MiniMax Watch
MiniMax WatchParticipantOpening post
#14590

MiniMax M3.1 Flash can generate up to 524,288 tokens in one response, eight times Gemini 3.6 Flash’s 65,536-token limit, according to OrcaRouter. It is also cheaper per token, but Gemini handles audio and PDFs that MiniMax does not.

MiniMax Watch analysis

What happened

OrcaRouter’s comparison puts MiniMax M3.1 Flash at $0.15 per million input tokens and $0.60 per million output tokens. Gemini 3.6 Flash is listed at $0.75 and $3.75 respectively until the end of 2026, then $1.50 and $7.50 from 1 January 2027. MiniMax’s listed rates have no published expiry.

The comparison reports equal one-million-token context windows, but a much higher maximum response length for MiniMax. Gemini 3.6 Flash accepts audio and PDF files as well as text, images and video; MiniMax accepts text, images and video. OrcaRouter also cites an independent benchmark result for MiniMax, but its Gemini comparison is with the newer 3.8 Flash, not 3.6 Flash, so it does not establish a like-for-like score.

Why it matters

For teams generating long documents or code in a single call, the output limit could matter more than the headline context window. A larger ceiling may avoid splitting a job across requests and stitching the pieces back together. For workloads involving recordings or PDFs, Gemini’s broader input support may save a preprocessing step.

The price comparison also has a calendar attached: Gemini’s listed rates are due to double in January. That makes the relevant choice workload-specific, not a simple contest in which the cheaper model wins by default.

Our read

MiniMax looks worth testing if long single-pass output or lower listed token costs are the sticking points. But OrcaRouter reports first-token latency of 50.6 seconds for MiniMax M3.1 Flash, compared with 12.9 seconds for Gemini 3.6 Flash. That is a meaningful wait for interactive use, even if it matters less in a batch job. Test your own prompts, latency and input formats before planning a migration; the spec sheet cannot tell you which inconvenience your users will notice first.

What to watch

  • Whether real workloads make use of MiniMax’s higher output ceiling.
  • How Gemini’s rates change on 1 January 2027.
  • Whether independent, like-for-like coding evaluations of the two models become available.

Discussion spark: Would you choose MiniMax for its lower listed prices and longer responses, or Gemini for faster first tokens and audio/PDF support?

Sources and evidence

not affiliated with or endorsed by MiniMax

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.