🔍 Read the full analysis: MiniMax M3.1 Flash: What Happened When We Tested It Against Open Models on ThorstenMeyerAI.com
Get smart everyday buys delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
MiniMax released M3.1 Flash on October 19, 2026, pricing it at $0.15 per million input tokens and $0.60 per million output tokens. An independent index rated it below several open-weight competitors, while a publisher’s 24-article newsroom test was designed to assess factual accuracy and writing quality; the supplied results do not include its scores. The model’s weights are expected October 23, but its license and parameter count have not been disclosed.
MiniMax released M3.1 Flash on October 19, offering it through its API, subscription and coding products, as well as OpenRouter. The company priced the model at $0.15 per million input tokens and $0.60 per million output tokens; an independent index gave it an Intelligence Index score of 43, below several open-weight rivals. A separate newsroom test covered 24 articles, but the supplied source does not give its scores.
MiniMax presents M3.1 Flash as a low-cost, fast model for agent work. The company says it is nearly three times faster than its M3 predecessor, reaches more than 200 tokens per second in everyday use and costs about 70% less per agent task. Those are company claims, not independently established findings. Artificial Analysis measured 120 output tokens per second and 1.08 seconds to the first token.
The model has a one-million-token context window, adjustable reasoning effort and a listed cached-input price of $0.015 per million tokens. OpenRouter offered it free for its first 72 hours. Its weights had not yet been released at the time of the October 20 report; MiniMax said they would arrive October 23. The company has not published an official parameter count or license.
MiniMax’s own benchmark charts show a mixed comparison. M3.1 Flash scored 62.2 on Automation-Bench, ahead of DeepSeek V4.1 Flash 1017 at 60.5. DeepSeek scored higher on DeepSWE, 77.0 to 73.7, and Terminal-Bench, 42.3 to 38.1. These results are self-reported, and the company selected the comparison models; the source says the results have not been independently reproduced.
Infographic 01 · MiniMax M3.1 Flash
A step behind on the index. Ahead of every open model in our newsroom.
An independent intelligence index puts MiniMax M3.1 Flash a step behind the open-weight leaders. Writing real news from real sources, it beat all of them.
Independent index
Artificial Analysis Intelligence Index, 10 evaluations. Bar length on a 0–50 scale.
Artificial Analysis lists M3.1 Flash as proprietary until its weights are released.
Our newsroom test
Mean score, English articles, 1–5 by GPT-6.1 Sol and Claude Opus 5.5 with the source in context. Bar length on a 0–5 scale.
In German, M3.1 Flash (3.36) also leads every open model, but GPT-6 Luna is clearly ahead (3.62).
Sources: Artificial Analysis, MiniMax-M3.1-Flash and open-weights models (20 Oct 2026); Thorsten Meyer AI newsroom benchmark, 12 sources × English + German, run 20 Oct 2026.
Powered by Thorsten Meyer AI
Price Gains Meet Uneven Test Results
The release gives developers another low-priced model option for agent and coding workflows, where repeated calls can make token costs consequential. But a lower price per token does not automatically mean a lower cost for a completed task. Artificial Analysis found that M3.1 Flash used 180 million tokens to complete its index evaluations, compared with a 99 million-token median. That is a raw comparison for the index test, not proof that every user’s tasks will consume more tokens.
Performance also depends on the work being measured. MiniMax’s charts put M3.1 Flash ahead on one automation benchmark, while DeepSeek led on the listed coding tests. Artificial Analysis placed M3.1 Flash at 43 on its Intelligence Index, behind several open-weight models. For buyers, the practical question is not just headline speed or price, but whether the model performs reliably on their own tasks at an acceptable total cost.
The newsroom test adds a different kind of evidence: whether a model can turn source material into accurate articles without inventing details. Its method is described, but its comparative scores are missing from the supplied source. Readers therefore cannot use this report to judge how M3.1 Flash performed against the other tested models on that task.
As an affiliate, we earn on qualifying purchases.
From Space Bunny to M3.1 Flash
Before the public release, the model operated anonymously on OpenRouter as Space Bunny Alpha. The source reports that it went live September 23, led OpenRouter’s usage ranking for 10 days and processed 56.7 trillion tokens there before being removed October 5. MiniMax says the model processed more than 200 trillion tokens overall; that broader figure is the company’s account.
The anonymous period also raises a data-handling issue for users who tried the service. According to the report, the provider could retain prompts during that period but was not allowed to train on them. MiniMax has not said what happened to the retained data. The report also notes an unconfirmed estimate of the model’s parameter count from another outlet, but excludes it because MiniMax has not verified the figure.
Artificial Analysis listed M3.1 Flash as proprietary at the time of measurement, because its weights were not yet available. That status could change after the promised release, but the eventual license will determine how developers may use, modify and redistribute the model.
“The company says M3.1 Flash is “nearly three times faster” than M3.”
— MiniMax
low-cost AI chatbot development tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Weights, Data Handling and Scores
Several points remain unresolved. The weights, license and official parameter count were not available in the source report, and the promised October 23 release had not occurred as of October 20. It is not yet clear whether the weights will be released on schedule or under terms that allow broad use and modification.
The fate of prompts retained during the Space Bunny Alpha period is also unknown. MiniMax has not explained whether those prompts were deleted, how long they were stored or whether users will receive further information. In addition, the company’s benchmark results have not been independently reproduced in the material provided. The newsroom test’s scores and comparison table are absent, so its outcome cannot be reported.
As an affiliate, we earn on qualifying purchases.
The October 23 Weights Release
The next stated milestone is MiniMax’s promised October 23 weights release. Its publication would allow observers to check the model’s architecture claims and assess the license and practical access terms, if those details are included. Independent benchmarks and real-world evaluations could then test whether its speed, quality and token use match the company’s positioning.
For users weighing the model now, the available evidence supports a cautious comparison: check performance on the specific workload, measure total tokens and latency, and review data-handling terms. The source report does not provide the newsroom benchmark’s final scores or confirm what MiniMax will say about Space Bunny prompts, leaving both issues open.
AI model with 24k token context window
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
When did MiniMax release M3.1 Flash?
MiniMax released M3.1 Flash on October 19, 2026, according to the source report.
How much does M3.1 Flash cost?
The listed price is $0.15 per million input tokens and $0.60 per million output tokens. Cached input is listed at $0.015 per million tokens.
How did it compare with DeepSeek V4.1 Flash?
In MiniMax’s charts, M3.1 Flash scored higher on Automation-Bench, while DeepSeek V4.1 Flash 1017 scored higher on DeepSWE and Terminal-Bench. The reported results are company-selected and self-reported.
Are M3.1 Flash’s weights available?
They had not been released by the October 20 report. MiniMax said the weights would arrive on October 23; the model’s license was not yet disclosed.
What happened to prompts sent to Space Bunny Alpha?
The report says the provider could retain prompts during the anonymous period but could not use them for training. MiniMax has not said what happened to the retained data.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
