Models / MiniMax
MiniMax-M3.1-Flash
by MiniMax
Used for coding, agents and long documents. Calls tools correctly on many tasks.
Best forCodingAgents and toolsLong documents
MiniMax-M3.1-Flash is made by MiniMax. It takes text and images. On Tīrtha it takes text.
Price per 1M tokens
The live price, read from billing as this page loads. You pay only for the tokens you use, from prepaid credit.
Limits and support
- Tool callingYesAlso measured on Tīrtha Bench
- Text inYes
- Image inYesText only on Tīrtha for now.
- StreamingYes
- Sampling settingsYou can set themThe maker suggests temperature 1, top p 0.95.
What it is for
- Coding
a native multimodal mixture-of-experts model for coding, agentic workflows, computer use, and cybersecurity.
From the maker's card - Agents and tools
for coding, agentic workflows, computer use, and cybersecurity.
From the maker's card - Long documents
with a context window of up to 1M tokens.
From the maker's card
In the maker's own words, from its published model card. Claims and benchmarks on that card are the maker's own, not measured by Tīrtha.
What a job costs
1,000 calls of 2,000 tokens in and 500 out cost about -
Worked out from the live price above, without cached input. Your own calls will differ in size.
Licence
Weights: MiniMax Community License. Read the licence.
Built with MiniMax M3.1-Flash
From the maker: Model card: MiniMaxAI/MiniMax-M3.1-Flash on Hugging Face · Last updated by the maker: Oct 26, 2026Facts from the maker's card are the maker's own, not measured by Tīrtha.
Tīrtha Bench review
What people use it for, in its maker's words:
What we found in our own tests:
- Agents and tool useCalls tools correctly on many tasks. Best for simpler agents with a short list of tools.
Our results: tool calling
Tool calling (public BFCL v4 test)
Overall: 84.9%*. Picks the right tool and fills in the inputs correctly on 85 of every 100 tasks.
One tool 84% · Picking the right tool 82% · Several calls at once 89%
A task counts as right only if the model calls the right tool with the right inputs. 800 tasks. Run on Oct 29, 2026. *Measured by Spanda Works on third-party hosts, not through Tīrtha. Results through Tīrtha may differ.
Handled several calls at once in our tests.
Similar models on Tīrtha
Try it
Model id catalog/minimax-m3.1-flash · base URL https://api.tirtha.ai/v1. Any OpenAI-compatible client works: change the base URL and keep your code.
Using a coding tool? Set up OpenCode, Cline, Cursor and more. PowerShell and the full reference: Quickstart.