Prototype · closed beta
unOpen.AI
In closed beta: Roxy, a private AI assistant with permanent memory running on infrastructure the user owns. Brand, site, product, backend and conversational design.
Model families
MiniMax.
Low prices, open weights.
MiniMax publishes weights for most of its language models and lists them at low prices. That makes it a candidate for high-volume, well-defined work, once a replay evaluation confirms the quality holds on your task.
All Model Families →
Current release
MiniMax today.
MiniMax ships these models as a range rather than one lead model, so this page is about the family: what it is good at, where it runs and how we evaluate it.
Last reviewed on 30 October 2026. We check the public model catalogues every week; when a new release appears, we review it against the provider's announcement and update this page.
LATEST MODELS
The newest MiniMax models.
What MiniMax has shipped most recently. Each has its own page with specifications, independent ratings at every reasoning setting and listed prices.
19 Oct 2026 · open weights
MiniMax M3.1 Flash
- Intelligence
- 28.9
- Context
- 1.05M
- In / out per M
- $0.075 / $0.300
1 Jun 2026 · open weights
MiniMax M3
- Intelligence
- 18.7
- Context
- 1.05M
- In / out per M
- $0.300 / $1.20
18 Mar 2026 · open weights
MiniMax M2.7
- Intelligence
- —
- Context
- 205k
- In / out per M
- $0.300 / $1.20
12 Feb 2026 · open weights
MiniMax M2.5
- Intelligence
- —
- Context
- 205k
- In / out per M
- $0.300 / $1.20
EVERY MODEL
All 8 MiniMax models, compared.
Sort by any column. Select a model for its full page.
| MiniMax M3.1 Flash open | 19 Oct 2026 | 28.9 | 111 | 1.05M | $0.075 | $0.300 |
|---|---|---|---|---|---|---|
| MiniMax M3 open | 1 Jun 2026 | 18.7 | 110 | 1.05M | $0.300 | $1.20 |
| MiniMax M2.7 open | 18 Mar 2026 | — | — | 205k | $0.300 | $1.20 |
| MiniMax M2.5 open | 12 Feb 2026 | — | — | 205k | $0.300 | $1.20 |
| MiniMax M2.1 open | 23 Dec 2025 | — | — | 205k | $0.300 | $1.20 |
| MiniMax M2 open | 26 Oct 2025 | — | — | 205k | $0.255 | $1.02 |
| MiniMax M1 | 17 Jun 2025 | — | — | 1M | $0.550 | $2.20 |
| MiniMax-01 open | 15 Jan 2025 | — | — | 1M | $0.200 | $1.10 |
Intelligence and coding: Artificial Analysis indices at the model's best reasoning setting, and its measured output speed. Context and prices: the OpenRouter catalogue list price in USD per million tokens. Released: the provider's date where Artificial Analysis records one, otherwise the date the catalogue first listed it. Refreshed weekly; not measured by QQuantum.ai. Select a heading to sort.
Specification
- Provider
- MiniMax, Shanghai, China
- Weights
- Published on Hugging Face (MiniMaxAI) for most language models in the catalogue
- Licence
- MiniMax M3.1 Flash carries the MiniMax community licence, not an OSI licence. Read the model card of the release you deploy
- Access paths
- MiniMax's API, third-party model catalogues, or an engineered deployment of the published weights
- Where we would test it
- High-volume extraction, classification and drafting where a low unit cost matters most
Provider-published specifications as of 30 October 2026, not figures we measured. Model families move quickly — check the provider's own documentation before committing to a number.
Licence: the MiniMax M3.1 Flash model card, checked 30 October 2026.
INDEPENDENT RATINGS
How MiniMax compares.
Third-party scores and listed prices, refreshed every week. Useful for building a shortlist; your own evaluation decides.
Composite index across their reasoning, knowledge, maths and coding evaluations · Higher is better · Linear scale from 0 to 66.0
- Gemini 4 Carbon
- 59.4
- Claude Opus 5.5
- 58.4
- GPT-6 Astra
- 56.9
- Muse Wonder Full
- 49.7
- MiniMax-M3.1-Flash
- 28.9
- MiniMax-M3
- 18.7
This family's highest-rated models, highlighted, against the best-rated model of each other family we cover. One composite score cannot show cost, latency, data residency or how a model handles your task.
Source: Artificial Analysis. Checked 30 October 2026. Measured independently by Artificial Analysis, not by QQuantum.ai or the providers.
USD per million output tokens · The newest releases in the family · Linear scale from 0 to $2.00
- MiniMax M3.1 Flash
- $0.30
- MiniMax M3
- $1.20
- MiniMax M2.7
- $1.20
- MiniMax M2.5
- $1.20
- MiniMax M2.1
- $1.20
Input is priced separately: MiniMax M3.1 Flash $0.08, MiniMax M3 $0.30, MiniMax M2.7 $0.30, MiniMax M2.5 $0.30, MiniMax M2.1 $0.30 per million tokens. Catalogue list prices; the provider's own rates, caching and batch discounts can differ.
Source: OpenRouter model catalogue. Checked 30 October 2026. Listed by OpenRouter, not measured by QQuantum.ai.
Where we would choose something else
A low price per token only matters if the answers pass. Compare cost per accepted result against a small model from a family you already use; on narrow tasks a fine-tuned small model can win on both quality and cost.
Where policy rules out models from Chinese providers, Mistral is the adjacent EU option. Where it does not, separate the hosted API, which sends data to the provider, from published weights on your own hardware, which do not.
Questions
Questions about MiniMax.
Is MiniMax open source?
Most of its language models publish weights, but under MiniMax's community licence rather than a standard open-source licence. Check the conditions against your intended use before deploying.
Why is it so cheap?
The listed prices are the provider's and catalogue's; they say nothing about quality on your work. A replay evaluation that counts accepted answers shows whether the saving survives.
What does the provenance review involve?
Pin the artefact, keep its licence, inspect the serving stack for outbound connections and test behaviour on your own topics. For the hosted API, review processing locations, retention and terms.
