Skip to main content

Vendor · Language models · Video models · Audio models

MiniMax Models: All 15 Models Compared

MiniMax has 15 models in the directory. The strongest is MiniMax M3: Rank 46 to 54 of 95 in the overall index.

Models159 of them since January 2026
Best rank46MiniMax M3, Rank 46 to 54 of 95
Cheapest API$0.15MiniMax M3.1 Flash, input per 1M tokens
Planfrom €11.18Hailuo AI Standard, per month incl. VAT, free plan available
Most popular66MiniMax H3, popularity index out of 100

Summary

In the directory: 11 language models, 2 video models and 2 audio models. 12 of them with open weights that can run locally.

All 15 models from MiniMax

Every model there ever was. Models with a profile lead to every benchmark, price and source.

15 models
Model type
Status
Weights

Language models · 11

MiniMax M3.1 FlashOct 19, 2026Activen/a$0.15 / $0.601181M47
MiniMax M3Jun 1, 2026Active12Rank 46 to 54 of 95$0.30 / $1.201071M54
MiniMax M2.7Mar 18, 2026Activebelow 1Rank 67 to 74 of 95$0.30 / $1.2035205K24
MiniMax-M2.7-highspeedMar 18, 2026Activen/an/an/a205K28
MiniMax-M2.5-highspeedFeb 13, 2026Activen/an/an/a205K13
MiniMax M2.5Feb 12, 2026Activebelow 1Rank 70 to 77 of 95$0.30 / $1.2038205K22
MiniMax M2.1Dec 20, 2025Activebelow 1Rank 72 to 82 of 95$0.30 / $1.2038205K31
MiniMax M2Oct 22, 2025Activen/a$0.26 / $1.02120205K38
MiniMax M1Jun 13, 2025Activen/a$0.40 / $2.20n/a1M27
MiniMax-01Jan 12, 2025Activen/a$0.20 / $1.10n/a1M35
MiniMax M3.1 Flash Previewn/aPreviewn/an/an/a1Mbelow 1

Video models · 2

MiniMax H3July 2026Active$7.80 per minute66
Hailuo 2.3Oct 28, 2025Active$4.90 per minute21

Audio models · 2

MiniMax-Music3Aug 13, 2026Open weightsn/a41
MiniMax Speech 2.8Jan 23, 2026Activen/abelow 1

The current MiniMax models explained

MiniMax M3.1 Flash

Language model · Oct 19, 2026

MiniMax M3.1 Flash has been available since October 19, 2026, through the API, M Plan, and MiniMax Code. Its official weights were publicly accessible on October 24. ModelScope history dates the first commit containing weight files to October 22; the date when they first became public is not established. The model card lists approximately 428 billion total parameters and 23 billion active parameters per token. This multimodal MoE model accepts text, images, and video, with a 1,000,000-token API context window. Its local-weight configuration sets max_position_embeddings to 1,048,576.

MiniMax-Music3

Audio model · Aug 13, 2026

MiniMax-Music3 generates complete songs from lyrics and a detailed music description. Its published card mentions vocals, accompaniment, editable musical structures, and 32 kHz stereo WAV output of up to five minutes.

MiniMax H3

Video model · July 2026

What is distinctive about MiniMax H3 is the combination of video and stereo audio in an open 33-billion-parameter model. It understands text, image, video, and audio references, supports first-and-last-frame workflows and multiple files, and generates 768p through H3-Base. 2K comes through H3-Regenerate-2K.

MiniMax M3

Language model · Jun 1, 2026

MiniMax M3 combines a 1M-token context with native image and video input plus computer use. Its distinctive feature is an interactive user simulator that reproduces requirement clarification, feedback, and project iteration for training and evaluation. Custom sparse attention supports these long coding and agent runs.

MiniMax M2.7

Language model · Mar 18, 2026

MiniMax M2.7 is the stable point release of the M2 line with an unusually concrete agent focus. MiniMax describes agent harnesses, Agent Teams, Skills, and dynamic tool search for software development, debugging, and long-running work. The openly available checkpoint keeps roughly 205K tokens of context, and an API is also available.

MiniMax-M2.7-highspeed

Language model · Mar 18, 2026

MiniMax-M2.7-highspeed is the faster API identifier for M2.7 serving. MiniMax describes M2.7 and M2.7-highspeed as two API versions with identical results but higher speed. MiniMax M2.7 supplies the documented capabilities, open weights, and agent focus.

MiniMax-M2.5-highspeed

Language model · Feb 13, 2026

MiniMax-M2.5-highspeed names the faster API serving path for MiniMax M2.5. MiniMax describes both versions as producing identical results, with Highspeed adding speed. The label provides no separate context or output specification.

MiniMax M2.5

Language model · Feb 12, 2026

MiniMax M2.5 was released for coding, search, office workflows, and training in real environments. The openly available MoE model offers roughly 205K tokens of context and is available through both an API and released weights.

MiniMax Speech 2.8

Audio model · Jan 23, 2026

MiniMax Speech 2.8 controls expressive speech synthesis through native sound tags and voice cloning from a ten-second reference. The family produces multilingual output and explicitly names Mandarin and Japanese among its supported languages.

MiniMax M2.1

Language model · Dec 20, 2025

MiniMax M2.1 is the direct update to MiniMax M2 and keeps its 230B MoE base with 10B active parameters and roughly 205K tokens of context. MiniMax names multilingual programming and complex work in real workflows as its target. Both weights and API access remain available.

Hailuo 2.3

Video model · Oct 28, 2025

New in Hailuo 2.3 are, according to MiniMax, more precise complex body movement, better response to motion commands, lighting and physics, and finer facial expressions. The general video model runs through API and web access and generates up to 1080p and 10 seconds.

MiniMax M2

Language model · Oct 22, 2025

MiniMax M2 opens the openly available M2 line with 230B total and 10B active parameters and roughly 205K tokens of context. MiniMax provides it both through an API and for custom inference.

2 more current models are in the model list above.

Which MiniMax model fits?

Derived from the overall index, the API price and the measured speed of the current models.

The strongest choice

MiniMax M3

The vendor's highest overall index

Overall index
12 (Rank 46 to 54 of 95)
API per 1M tokens
$0.30 / $1.20
Speed
107 tokens/s

The fastest

MiniMax M2

120 tokens/s, a typical model manages 98

API per 1M tokens
$0.26 / $1.02
Speed
120 tokens/s

MiniMax in the overall index

The 20 best language models, with the models from MiniMax highlighted. “All” and the filter show every model in the index.

Overall index (Gradually Meta-Score v3)

20 of 95 models

Claude Opus 5.596.5
Gemini 4 Carbon96.1
Claude Fable 5.195.2
Claude Sonnet 5.593.9
Gemini 4 Argon93.5
GPT-6 Astra92.0
GPT-6.1 Sol90.0
Muse Wonder Full89.8
Claude Opus 587.8
Muse Wonder82.5
Kimi K3.1 Pro79.9
Claude Fable 579.8
GPT-6 Sol73.4
GPT-5.6 Sol71.3
Gemini 3.7 Flash69.6
Muse Spark 1.369.4
Gemini 3.8 Flash67.3
DeepSeek V4.1 Flash 101765.9
GPT-6.1 Terra64.2
MiniMax M312.0

Where does MiniMax stand?

Models among the best 20 in the overall index

Anthropic5 models, 4 of them among the best ten
OpenAI5 models, 2 of them among the best ten
Google4 models, 2 of them among the best ten
Meta3 models, 2 of them among the best ten
DeepSeek1 models, 0 of them among the best ten
Moonshot AI1 models, 0 of them among the best ten
xAI1 models, 0 of them among the best ten
MiniMax0 models, 0 of them among the best ten

Best MiniMax model per area

Agents
Rank 13 to 37 of 82
MiniMax M3.1 Flash
Health
Rank 13 to 66 of 69
MiniMax M2.1
Knowledge
Rank 15 to 25 of 26
MiniMax M3
Reasoning
Rank 20 to 30 of 87
MiniMax M3.1 Flash
Coding
Rank 23 to 49 of 89
MiniMax M3.1 Flash
General
Rank 23 to 42 of 56
MiniMax M3
Multimodal
Rank 25 to 31 of 47
MiniMax M3
Legal
Rank 29 to 56 of 67
MiniMax M3
Mathematics
Rank 34 to 39 of 44
MiniMax M2.1
Finance
Rank 44 to 55 of 97
MiniMax M3

Vendors compared

VendorModelsModel typesBest language modelTop 10Open weightsNew in 2026
Anthropic3333 language modelsClaude Opus 5.5 (Rank 1 to 3)4013
OpenAI6346 language models, 4 image models, 2 video models, 9 audio models, 2 specialist modelsGPT-6 Astra (Rank 4 to 8)2427
Google6643 language models, 4 image models, 6 video models, 11 audio models, 2 specialist modelsGemini 4 Carbon (Rank 1 to 5)21926
Meta3229 language models, 1 image model, 1 video model, 1 specialist modelMuse Wonder Full (Rank 6 to 9)2267
DeepSeek2323 language modelsDeepSeek V4.1 Flash 1017 (Rank 13 to 20)0239
Moonshot AI1212 language modelsKimi K3.1 Pro (Rank 10 to 13)0107
xAI1612 language models, 2 image models, 2 video modelsGrok 4.7 (Rank 18 to 24)0112
MiniMax1511 language models, 2 video models, 2 audio modelsMiniMax M3 (Rank 46 to 54)0129

Strengths

Rank 46 to 54 of 95 in the overall index (MiniMax M3)
MiniMax M3.1 Flash from $0.15 per 1M input tokens
12 models with open weights

Gaps

No image models in the directory
1 models only with limited access or as a preview

Every MiniMax model over time

By release date. Grey: retired models.

20255 models

JanMiniMax-01
JunMiniMax M1
OctMiniMax M2
OctHailuo 2.3Video model
DecMiniMax M2.1

20269 models

JanMiniMax Speech 2.8Audio model
FebMiniMax M2.5
FebMiniMax-M2.5-highspeed
MarMiniMax M2.7
MarMiniMax-M2.7-highspeed
JunMiniMax M3
JulMiniMax H3Video model
AugMiniMax-Music3Audio model
OctMiniMax M3.1 Flash

No documented date1 model

MiniMax M3.1 Flash Preview

Releases in the last twelve months

1Dec 25
1Jan 26
2Feb 26
2Mar 26
0Apr 26
0May 26
1Jun 26
1Jul 26
1Aug 26
0Sep 26
1Oct 26
0Nov 26

9 releases since January 2026, by release date.

API prices

Prices per 1M tokens for the current language models, according to MiniMax or the provider linked in each model profile.

Current modelInputOutputContext
MiniMax M3$0.30$1.201M
MiniMax M2.7$0.30$1.20205K
MiniMax M2.5$0.30$1.20205K
MiniMax M2.1$0.30$1.20205K
MiniMax M3.1 Flash$0.15$0.601M
MiniMax-01$0.20$1.101M
MiniMax M2$0.26$1.02205K
MiniMax M1$0.40$2.201M

Image and video model prices

Current modelPrice
MiniMax H3$7.80 per minute
Hailuo 2.3$4.90 per minute

Plans from MiniMax

Monthly prices in Germany incl. VAT; “≈” is converted from US dollars. The vendor's models are often also part of other providers' plans, see the AI subscription comparison.

PlanPer monthNote
Hailuo AI Freefreewith credits
Hailuo AI Standard≈ €11.18with credits
MiniMax Token Plan Plus≈ €23.43
Hailuo AI Pro≈ €40.48with credits
MiniMax Token Plan Max≈ €58.58
Hailuo AI Master≈ €94.80with credits
MiniMax Token Plan Ultra≈ €140.60
Hailuo AI Max≈ €230.08with credits

Frequently asked questions about MiniMax

What MiniMax models are there?

The directory lists 15 MiniMax models: 11 language models, 2 video models and 2 audio models. It counts every model MiniMax ever released, including retired ones. 14 are currently available, the newest being MiniMax M3.1 Flash, MiniMax-Music3 and MiniMax H3.

Which MiniMax model is the best?

MiniMax M3. It is at rank 46 to 54 of 95 in the overall index, which combines independent benchmarks from several sources.

What is the latest MiniMax model?

MiniMax M3.1 Flash, released on Oct 19, 2026. The timeline below shows every release by date.

How much do MiniMax models cost?

Through the API per 1M tokens (input / output): MiniMax M3.1 Flash $0.15 / $0.60 and MiniMax M3 $0.30 / $1.20. The prices of all 8 current language models are in the price table. As a subscription, MiniMax models start at €11.18 per month incl. VAT (Hailuo AI Standard).

Can I use MiniMax models for free?

Yes, with the Hailuo AI Free plan. The cheapest paid plan is Hailuo AI Standard at €11.18 per month incl. VAT.

Is MiniMax better than ChatGPT?

In the overall index, no. The best OpenAI model, GPT-6 Astra (rank 4 to 8 of 95), is ahead of MiniMax M3 (rank 46 to 54 of 95). Single tasks can differ.

Can I run MiniMax models locally?

Yes, 12 of the 15 models have open weights. Each model profile lists how much memory it needs.

Which MiniMax model is the fastest?

MiniMax M2 with 120 tokens per second according to Artificial Analysis. A typical model manages 98.

How large is the context window of the MiniMax models?

Up to 1 M tokens (≈ 1,500 A4 pages), which 5 of the current models reach, including MiniMax M3.1 Flash. Every model's value is in the model list.

Sources and data date

Data as of Nov 1, 2026

Show all 10 sources
MiniMax M3.1 Flash: API prices
MiniMax M3: API prices
MiniMax M2.7: API prices
MiniMax M2.5: API prices
MiniMax M2.1: API prices
MiniMax M2: API prices
MiniMax M1: API prices
MiniMax-01: API prices
Artificial Analysis: output speed
AI subscription comparison (plans and sources)