Skip to main content

MiniMax · Text

MiniMax models: Versions, specs, GPU requirements & hosting

MiniMax M-series models focus on efficient sparse reasoning and agentic deployment across a clearly ordered general track.

Current releases Browse engineering services Compare models
Technical deployment reference illustration for MiniMax
Reference deployment flow; exact architecture facts are listed separately.

Overview and history

MiniMax is an official MiniMax model family represented here through independently verified release, checkpoint, license, runtime and hardware evidence. MiniMax M-series models focus on efficient sparse reasoning and agentic deployment across a clearly ordered general track. Current public releases are grouped by the General tracks so sibling sizes and specialist branches are not presented as false chronological generations.

FusionPointAI currently tracks 10 certified public MiniMax releases. The catalogue preserves release dates and evaluates Current, Recent, Previous and Legacy status within each model track rather than by database insertion order.

Sparse MoE efficiency
Reasoning workflows
Long-context operation

Architecture and model range

The verified public range spans 229B to 456B total parameters and context windows up to 10,240,000 tokens. Each release states whether it is dense or mixture-of-experts and records active parameters separately when the official source publishes them.

Public releases10
Model tracks1
Current generationMiniMax M3
Last verified2026-09-03T00:00:00.000000Z

Generations, tracks and variants

General

MiniMax M2

Legacy generation
MiniMax M2

230.0B parameters · 196,608 context

MoE MiniMaxM2ForCausalLM

MiniMax M1

Legacy generation
MiniMax M1 80K

456.0B parameters · 10,240,000 context

Reasoning · MoE MiniMaxM1ForCausalLM

MiniMax M2.1

Legacy generation
MiniMax M2.1

230.0B parameters · 196,608 context

MoE MiniMaxM2ForCausalLM

MiniMax M2.5

Previous generation
MiniMax M2.5

230.0B parameters · 196,608 context

MoE MiniMaxM2ForCausalLM

minimax-m3-dspark

Recent generation
MiniMax-M3-DSpark

The manufacturer has not published this information. parameters · N/A / Not directly comparable context

Fine-tune Unknown / Not Recommendable

MiniMax M2.7

Recent generation
MiniMax M2.7

229.0B parameters · 204,800 context

MoE MiniMaxM2ForCausalLM Separate Commercial License Required

minimax-m3

Current generation 4 sibling variants
MiniMax-M3-NVFP4

The manufacturer has not published this information. parameters · N/A / Not directly comparable context

Quantization Unknown / Not Recommendable
MiniMax-M3

The manufacturer has not published this information. parameters · N/A / Not directly comparable context

Base Unknown / Not Recommendable
MiniMax-M3-MXFP8

The manufacturer has not published this information. parameters · N/A / Not directly comparable context

Quantization Unknown / Not Recommendable
MiniMax M3

428.0B parameters · 1,048,576 context

MoE MiniMaxM3SparseForConditionalGeneration Unknown / Not Recommendable

Deployment ecosystem

Choose a specific MiniMax release and certified checkpoint before sizing hardware. FusionPointAI rejects unknown runtime configurations, calculates resident memory from total model weights, then distinguishes globally compatible GPUs from currently rentable inventory.

Validate the exact checkpoint license, runtime version, precision kernel, context length, concurrency and multi-GPU topology. Active MoE parameters affect compute, but resident model memory continues to use total stored weights.

Common use cases

Agent backends
Reasoning services
Long-form generation

License overview

Licenses are certified per release, not inferred from the provider. Review the release page before commercial deployment because terms can differ between generations and variants.

Commercial-use status is a source-backed catalogue classification, not legal advice. Follow the linked official license and attribution requirements.

Release timeline

minimax/minimax/minimax-m1-80k

Source details

Official source

Source revision: 5e1291d5b7dfe2a74150c3eeeb6c6feba553622c9e2983615496c5b280cd286c

minimax/minimax/minimax-m3-nvfp4

Source details

Official source

Source revision: e125d29d880a335177c8afd6f0c9af13d29615293f674c10343c2c21416253db

minimax/minimax/minimax-m3-mxfp8

Source details

Official source

Source revision: 3862c1f28b912545bce9121f0bfe5b689efbff984373e1b16ca071a6292bb3d4

minimax/minimax/minimax-m3-dspark

Source details

Official source

Source revision: 292a907931089d171aab35c37879f29b5fcb282cb01a1e5af3695a1187ea8f3d

minimax/minimax/minimax-m3

Source details

Official source

Source revision: 379adf789f3179a77837f6818a8b981c05678577dd940948d3bfc980bc622f4d

minimax/minimax/minimax-m2-7

Source details

Official source

Source revision: c138b20640417457b4614efc2bd484f84839f87024c40db6defde246235ebe0b

minimax/minimax/minimax-minimax-m3

Source details

Official source

Source revision: 36685ff255b473e07e602b2781872b6ab5b68c6a64838c8b9a2165cc5c5129b5

minimax/minimax/minimax-m2-5

Source details

Official source

Source revision: dcbb3c5c80486ed208d15ef8fdc522a20eb11d69fb8018deef95d37452923d71

minimax/minimax/minimax-m2-1

Source details

Official source

Source revision: a51f01f8e16a8ef0dc2eaab0907c17ba0debd56c159085ff05e49b36e7b8d58c

minimax/minimax/minimax-m2

Source details

Official source

Source revision: 1033584f50c45eb3dec4ad9af87212ce2f76fd319de338bfcfa3c561ea8b0e24

MiniMax/MiniMax-M3 @ 2026-07-23T13:56:57Z

Source details

Official source

Source revision: 2026-07-23T13:56:57Z

Frequently asked questions

Which MiniMax release is current?

Current releases are MiniMax M3, MiniMax-M3, evaluated within their respective tracks.

Can MiniMax be self-hosted?

Only releases with a verified official checkpoint and supported runtime configuration are presented as deployable.

Which GPU should run MiniMax?

GPU selection depends on the exact variant, checkpoint precision, context length and concurrency; use the release hardware recommendations rather than the family name alone.

Official references

Last verified: 2026-09-03