🔍 Read the full analysis: MiniMax M3.1 Flash: Examining Its Leaderboard Rank And Test Performance on ThorstenMeyerAI.com
Get the latest gadgets 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. Artificial Analysis gives it an Intelligence Index score of 43, behind several open-weight competitors, while a separate newsroom test is described but its results are missing from the supplied source material. The model’s weights are promised for October 23.
MiniMax released M3.1 Flash on October 19, pricing the model at $0.15 per million input tokens and $0.60 per million output tokens. The release identifies the model that had operated anonymously as Space Bunny Alpha on OpenRouter; independent testing gives it a score of 43 on Artificial Analysis’ Intelligence Index, below several open-weight rivals, while MiniMax’s own benchmarks show a mixed comparison.
MiniMax says M3.1 Flash is designed for fast, inexpensive agent work and costs half as much per token as its predecessor, M3. Its published rates also include $0.015 per million cached input tokens. OpenRouter made the model available free for its first 72 hours. MiniMax claims typical speeds above 200 tokens per second, with a median above 170, and says it is nearly three times faster than M3. Those speed figures are company claims, not the independent measurement.
Artificial Analysis measured 120 tokens per second and 1.08 seconds to the first token. Its Intelligence Index score for M3.1 Flash is 43. That trails MiMo-V2.6-Pro at 46, DeepSeek V4.1 Flash 1017 and GLM-5.3 at 45, and Step 5 Preview and Kimi K3 at 44. Artificial Analysis also reports that M3.1 Flash used 180 million tokens to complete its index evaluation, compared with a 99 million-token median. That is relevant to total usage costs: a low token price may be offset if a model consumes more tokens to finish work.
MiniMax’s published comparison shows M3.1 Flash ahead on Automation-Bench, scoring 62.2 against DeepSeek V4.1 Flash 1017’s 60.5. DeepSeek leads on DeepSWE, 77.0 to 73.7, and Terminal-Bench, 42.3 to 38.1. On OSWorld V2 and Agents’ Last Exam, MiniMax’s chart does not report DeepSeek scores. The source material describes a separate newsroom test of 24 articles, but its score table and results are absent, so no outcome from that test can be reported here.
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, Speed and Real-World Results
M3.1 Flash’s low listed price could make it attractive for developers running repeated model calls, including agent workflows. But a per-token price is not the same as the cost of completing a task. The independent index’s token-use measurement suggests buyers should compare total tokens consumed, speed and output quality alongside the advertised rates.
The gap between MiniMax’s speed claims and Artificial Analysis’ measurement also matters for teams weighing latency. The two figures may reflect different test conditions, and the supplied information does not establish why they differ. Until the model is tested under comparable workloads and the weights become available, the evidence supports a narrower conclusion: M3.1 Flash is competitively priced, but it is not at the top of the cited index rankings.
As an affiliate, we earn on qualifying purchases.
From Space Bunny to M3.1 Flash
Before the official release, the model appeared on OpenRouter as Space Bunny Alpha. According to the source report, it went live anonymously on September 23, led OpenRouter’s usage ranking for 10 days and processed 56.7 trillion tokens on that platform before being removed on October 5. MiniMax says it processed more than 200 trillion tokens overall; the report does not provide independent verification of that company-wide figure.
MiniMax’s benchmark comparisons were published as chart images and transcribed by CellCog, according to the source report. The company selected the comparison models, and the chart includes some older versions of competing systems. The report notes that MiniMax’s “first among Chinese models” claim on Agents’ Last Exam needs qualification: DeepSeek’s model card lists a score of 35.4 for V4.1 Flash 1017, while MiniMax’s chart does not show a DeepSeek result for that benchmark. Differences in test setup may affect the comparison.
““nearly three times faster” than M3”
— MiniMax
affordable AI model for agent tasks
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Weights, Data and Benchmark Gaps
The model weights were not yet available in the supplied report, and MiniMax had promised them for October 23. No license or official parameter count was provided. A separate outlet reportedly estimated 428 billion total parameters and 23 billion active parameters, but MiniMax had not confirmed those figures, so they should not be treated as established specifications.
The source report says that during the anonymous Space Bunny period, the provider could retain prompts but could not use them to train the model. MiniMax had not said what happened to those stored prompts. The report also states that MiniMax’s benchmark figures, including its security evaluation, had not been independently reproduced. The newsroom test’s scoring results are missing from the provided material, leaving its stated 24-article setup without a reportable conclusion.
As an affiliate, we earn on qualifying purchases.
The October 23 Weights Release
The next stated milestone is MiniMax’s planned October 23 weights release. Users and independent evaluators can then check whether the release occurs, review the licensing terms and assess the model beyond API access. The supplied material does not confirm a license or say whether the promised release will include enough documentation to establish the architecture and parameter count.
Further independent comparisons could clarify the differences between MiniMax’s speed claims and Artificial Analysis’ measurements, as well as how the model performs on practical tasks and how much it costs to complete them. MiniMax has not provided an explanation for the Space Bunny prompt-retention question in the source report. Those details, along with the missing newsroom-test scores, remain open.
As an affiliate, we earn on qualifying purchases.
Key Questions
What is MiniMax M3.1 Flash?
M3.1 Flash is a model released by MiniMax on October 19, 2026, marketed for fast, lower-cost agent work. It was previously available anonymously on OpenRouter as Space Bunny Alpha, according to the source report.
How much does M3.1 Flash cost?
MiniMax lists the price at $0.15 per million input tokens, $0.60 per million output tokens and $0.015 per million cached input tokens. Actual task costs also depend on how many tokens a model uses.
How did it score against competitors?
Artificial Analysis gives M3.1 Flash an Intelligence Index score of 43, below several listed open-weight models. MiniMax’s own charts show it leading on Automation-Bench, while DeepSeek V4.1 Flash 1017 scores higher on DeepSWE and Terminal-Bench.
Are the model weights available?
Not according to the source report dated October 20. MiniMax said the weights would arrive October 23, but the report does not confirm their release or provide a license.
What happened to prompts sent to Space Bunny Alpha?
The source report says the provider could retain prompts during the anonymous period but could not use them for training. MiniMax had not explained what happened to that data by the report’s publication.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
