MiniMax M3.1 Flash
minimax/minimax-m3.1-flash- MiniMax
- Family · MiniMax M3 Flash
- Released 19 Oct 2026
- Open weights
- Text · Image · Video → Text
- Reasoning · adjustable
- Tools · JSON · Vision
Not on OSMEXO yet
Spec & list price
Context1M
Max output512K
Input / 1M$0.15
Output / 1M$0.604× input
Scores · standing
Intelligence42.7#27/489
LiveBench66.9#16/32
Usage
Tokens / day900.3B▲ 651% 7d
Usage rank#4
Requests / day6.6M
Tool calls / req0.61
Per request · 30 days
Avg input90.3K6.1× median
Avg output1.3K1.9× median
Cache hit93%1.3× median
Reasoning share68%1.2× median
Tool-call failures2.9%1.5× median
Consumption
Peak 1.1T 30 OctLow 121.1M 19 OctRank today #4
News & trends
57 mentions this week ▲ 90%this model · videos, posts, articles · 30 days
Trending · 14 days
- @ArtificialAnlys on X151.5K views
- MiniMax M3.1 Flash: una suscripción barata que rinde para programar con IA27.4K views
- I Tested MiniMax M3.1 Flash Inside Unreal Engine 512.4K views
- Run MiniMax-M3.1-Flash Locally Guide!189 upvotes
- MiniMax-M3.1-Flash369 upvotes
- @MiniMax_AI on X420.4K views
News · latest from MiniMax
Benchmarks
bar = share of the top score · Benchmark data: Artificial AnalysisSpeed · median across 14 providers, latest day
Output speed#82/33498 tok/s1.5× faster than median
Fastest provider200 tok/sbest of 14 providers
First token#1500.95 sabout median
First answer token#6218 sthinking included
#4 cheapest of 14 with a similar score
Value
Intelligence Index
Blended price per 1M tokens (log)
LiveBench suites
Browser80.6
Coding83.9
Data Analysis71.4
Design74.0
Long Agentic28.8
Reasoning82.9
Research72.7
▲ Strongest: Coding, Reasoning. ▼ Weakest: Long Agentic (29).
Technical details
- Reasoning
- on · max
- Tokenizer
- Other
- License
- other
- Parameters
- 220B
- Moderation
- none
- Downloads · likes
- 23.5K · 529
Reasoning effort
- max · default
- xhigh
- high
- medium
- low
Request parameters 20
- tools
- tool_choice
- response_format
- structured_outputs
- reasoning
- reasoning_effort
- include_reasoning
- temperature
- top_p
- top_k
+10 more
- min_p
- max_tokens
- stop
- seed
- frequency_penalty
- logit_bias
- logprobs
- presence_penalty
- repetition_penalty
- top_logprobs
Interesting facts
- 2× the output room of similar models (524.3K against a 262.1K median).
- Released 14 days ago.
- Weekdays average 462.8B tokens a day against 521.2B on weekends.
When to use
Best for
- Cost and latency sensitive coding help
- Image and video understanding in apps
- Long document analysis with adjustable reasoning
- Tool-calling assistants at steady volume
- Self-hosted deployments on open weights
Think twice for
- Long multi-step agent runs
- Hardest research-grade reasoning
- Image or video generation
Alternatives
other vendors · similar capability- 45.1
- 38.1
- 37.9
- 44.8
Want MiniMax M3.1 Flash on OSMEXO?
Sign in to request it. Requests are reviewed; a request does not guarantee availability.











