MiniMax model product
MiniMax M3
MiniMax M3 is the first open-weight model to combine three frontier capabilities: top-tier coding and agentic performance, a 1M-token context window, and native multimodality.
Updated Aug 17, 2026. Default version: MiniMax M3
LLMBoard Score71.1MiniMax M3
Coverage40%21 benchmark families
Context window512KTokens
Official input price$0.30MiniMax (minimax.io) API
MiniMax M3 Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
MiniMax M3 Benchmark Results
Benchmark scores for MiniMax M3.
Sort by
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkBankerToolBench | Score76.1% | Rank01 | Participants1 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkDRACO | Score73.2% | Rank01 | Participants1 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkGDPval-Rubrics | Score74.8% | Rank01 | Participants1 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkIMO 2025 | Score35 points | Rank01 | Participants2 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkKernelBench Hard | Score28.8% | Rank01 | Participants1 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkLiveSQLBench | Score40.2% | Rank01 | Participants1 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkLOCA-Bench (256k) | Score49.3% | Rank01 | Participants1 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkOmniDocBench 1.5 | Score91.6% | Rank01 | Participants18 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkSpreadSheetBench-v1 | Score89.3% | Rank01 | Participants3 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkSVG-Bench | Score63.7% | Rank01 | Participants1 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkSWE Atlas - Test Writing | Score30.8% | Rank01 | Participants1 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkSWE-fficiency | Score34.8% | Rank01 | Participants1 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkUSAMO 2026 | Score36 points | Rank01 | Participants3 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkVIBE-V2 | Score50.1% | Rank01 | Participants1 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkYC-Bench | Score2,100,000 points | Rank01 | Participants1 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkCL-bench | Score20.5% | Rank02 | Participants2 | Percentile0.0% | EvidenceC | Evaluated |
| BenchmarkPostTrainBench | Score37.1% | Rank02 | Participants6 | Percentile80.0% | EvidenceC | Evaluated |
| BenchmarkSWE Atlas - Codebase QnA | Score37.9% | Rank02 | Participants2 | Percentile0.0% | EvidenceC | Evaluated |
| BenchmarkClaw-Eval | Score74.5% | Rank03 | Participants14 | Percentile84.6% | EvidenceC | Evaluated |
| BenchmarkPaperBench | Score52.6% | Rank03 | Participants3 | Percentile0.0% | EvidenceC | Evaluated |
| BenchmarkVideo-MME | Score85.4% | Rank06 | Participants17 | Percentile68.8% | EvidenceC | Evaluated |
| BenchmarkAPEX-Agents | Score27.7% | Rank07 | Participants8 | Percentile14.3% | EvidenceC | Evaluated |
| BenchmarkOfficeQA Pro | Score45.1% | Rank07 | Participants7 | Percentile0.0% | EvidenceC | Evaluated |
| BenchmarkVideoMMMU | Score84.6% | Rank08 | Participants26 | Percentile72.0% | EvidenceC | Evaluated |
| BenchmarkFinance Agent v2 | Score48.3% | Rank09 | Participants26 | Percentile68.0% | EvidenceB | Evaluated |
| BenchmarkSWE-Bench Verified | Score80.5% | Rank10 | Participants111 | Percentile91.8% | EvidenceC | Evaluated |
| BenchmarkNL2Repo | Score42.1% | Rank12 | Participants17 | Percentile31.3% | EvidenceC | Evaluated |
| BenchmarkFrontierCode 1.1 | Score14.7% | Rank15 | Participants16 | Percentile6.7% | EvidenceB | Evaluated |
| BenchmarkMCP Atlas | Score74.2% | Rank15 | Participants33 | Percentile56.3% | EvidenceC | Evaluated |
| BenchmarkBrowseComp | Score83.5% | Rank16 | Participants62 | Percentile75.4% | EvidenceC | Evaluated |
| BenchmarkSWE-Bench Pro | Score59.0% | Rank16 | Participants50 | Percentile69.4% | EvidenceC | Evaluated |
| BenchmarkOSWorld-Verified | Score70.1% | Rank17 | Participants24 | Percentile30.4% | EvidenceC | Evaluated |
| BenchmarkMMMU-Pro | Score78.1% | Rank21 | Participants68 | Percentile70.2% | EvidenceC | Evaluated |
| BenchmarkTerminal-Bench 2.1 | Score66.0% | Rank21 | Participants28 | Percentile25.9% | EvidenceC | Evaluated |
| BenchmarkLiveBench | Score70.0% | Rank31 | Participants38 | Percentile18.9% | EvidenceB | Evaluated |
MiniMax M3 Arena Results
Preference and agent-evaluation results for the default version.
Sort by
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenaagent bash recovery steps | Categoryoverall | Rank24 | Rating / score0.1 | VotesN/A | Observations57.2K | Result date |
| Arenadocument | Categoryoverall | Rank26 | Rating / score1434.8 | Votes6,263 | ObservationsN/A | Result date |
| Arenaagent tool hallucination | Categoryoverall | Rank30 | Rating / score0.0 | VotesN/A | Observations1.9M | Result date |
| Arenadocument style control | Categoryoverall | Rank32 | Rating / score1433.5 | Votes6,263 | ObservationsN/A | Result date |
| Arenawebdev | Categoryoverall | Rank34 | Rating / score1489.8 | Votes10,053 | ObservationsN/A | Result date |
| Arenaagent | Categoryoverall | Rank37 | Rating / score-0.0 | VotesN/A | Observations2M | Result date |
| Arenaagent praise complaint | Categoryoverall | Rank37 | Rating / score-0.1 | VotesN/A | Observations7.9K | Result date |
| Arenaagent steerability | Categoryoverall | Rank37 | Rating / score-0.1 | VotesN/A | Observations31.4K | Result date |
| Arenaagent task outcome explicit | Categoryoverall | Rank39 | Rating / score-0.1 | VotesN/A | Observations19.9K | Result date |
| Arenavision | Categoryoverall | Rank42 | Rating / score1257.7 | Votes10,984 | ObservationsN/A | Result date |
| Arenavision style control | Categoryoverall | Rank47 | Rating / score1239.7 | Votes10,984 | ObservationsN/A | Result date |
| Arenatext | Categoryoverall | Rank71 | Rating / score1435.4 | Votes37,180 | ObservationsN/A | Result date |
| Arenatext style control | Categoryoverall | Rank72 | Rating / score1443.4 | Votes37,180 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryoverall | Rank84 | Rating / score1435.9 | Votes37,162 | ObservationsN/A | Result date |
MiniMax M3 Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $0.30 input, $1.2 output per 1M
- Official provider
- MiniMax (minimax.io)
- Lowest third-party
- From $0.225 input, $0.90 output per 1M via EmpirioLabs AI
- Tracked offerings
- 36
Sort by
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderNvidia | Provider model IDminimaxai/minimax-m3 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context1M | Updated |
| ProviderKenari | Provider model IDminimax-m3 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context512K | Updated |
| ProviderMiniMax Token Plan (minimaxi.com) | Provider model IDMiniMax-M3 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context1M | Updated |
| ProviderMiniMax Token Plan (minimax.io) | Provider model IDMiniMax-M3 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context1M | Updated |
| ProviderSCNet Token Plan | Provider model IDMiniMax-M3 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context512K | Updated |
| ProviderEmpirioLabs AI | Provider model IDminimax-m3 | Regionglobal | Input / 1M$0.225 | Output / 1M$0.90 | Context1M | Updated |
| ProviderWeights & Biases | Provider model IDMiniMaxAI/MiniMax-M3 | Regionglobal | Input / 1M$0.23 | Output / 1M$0.96 | Context262.1K | Updated |
| ProviderDeep Infra | Provider model IDMiniMaxAI/MiniMax-M3 | Regionglobal | Input / 1M$0.28 | Output / 1M$1.1 | Context524.3K | Updated |
| ProviderLilac | Provider model IDminimaxai/minimax-m3 | Regionglobal | Input / 1M$0.28 | Output / 1M$1.1 | Context1M | Updated |
| ProviderNanoGPT | Provider model IDminimax/minimax-m3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context512K | Updated |
| ProviderHugging Face | Provider model IDMiniMaxAI/MiniMax-M3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context524.3K | Updated |
| ProviderOpenRouter | Provider model IDminimax/minimax-m3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context1M | Updated |
| ProviderOpenCode Go | Provider model IDminimax-m3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context1M | Updated |
| ProviderNebius Token Factory | Provider model IDMiniMaxAI/MiniMax-M3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context1M | Updated |
| ProviderMerge Gateway | Provider model IDminimax/minimax-m3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context1M | Updated |
| ProviderTogether AI | Provider model IDMiniMaxAI/MiniMax-M3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context524.3K | Updated |
| ProviderOpenCode Zen | Provider model IDminimax-m3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context512K | Updated |
| ProviderMiniMax (minimaxi.com) | Provider model IDMiniMax-M3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context1M | Updated |
| ProviderClinePass | Provider model IDcline-pass/minimax-m3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context512K | Updated |
| ProviderVercel AI Gateway | Provider model IDminimax/minimax-m3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context1M | Updated |
| ProviderEden AI | Provider model IDminimax/MiniMax-M3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context524.3K | Updated |
| ProviderLLM Gateway | Provider model IDminimax-m3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context1M | Updated |
| ProviderKilo Gateway | Provider model IDminimax/minimax-m3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context524.3K | Updated |
| ProviderPioneer | Provider model IDMiniMaxAI/MiniMax-M3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context1M | Updated |
| ProviderFireworks AI | Provider model IDaccounts/fireworks/models/minimax-m3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context512K | Updated |
| ProviderAbacus | Provider model IDMiniMaxAI/MiniMax-M3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context1M | Updated |
| ProviderMiniMax (minimax.io) | Provider model IDMiniMax-M3 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context1M | Updated |
| ProviderCharm Hyper | Provider model IDminimax-m3 | Regionglobal | Input / 1M$0.3266 | Output / 1M$1.31 | Context512K | Updated |
| ProviderCrossModel | Provider model IDminimax/minimax-m3 | Regionglobal | Input / 1M$0.33 | Output / 1M$1.32 | Context1M | Updated |
| ProviderWafer | Provider model IDMiniMax-M3 | Regionglobal | Input / 1M$0.33 | Output / 1M$1.32 | Context1M | Updated |
| ProviderCortecs | Provider model IDminimax-m3 | Regionglobal | Input / 1M$0.395 | Output / 1M$1.98 | Context1M | Updated |
| ProviderTensorX | Provider model IDminimax/minimax-m3 | Regionglobal | Input / 1M$0.40 | Output / 1M$2 | Context1M | Updated |
| ProviderSynthetic | Provider model IDhf:MiniMaxAI/MiniMax-M3 | Regionglobal | Input / 1M$0.60 | Output / 1M$1.2 | Context524.3K | Updated |
| ProviderZenMux | Provider model IDminimax/minimax-m3 | Regionglobal | Input / 1M$0.60 | Output / 1M$2.4 | Context512K | Updated |
| ProviderOfox | Provider model IDminimax/minimax-m3 | Regionglobal | Input / 1M$0.60 | Output / 1M$2.4 | Context512K | Updated |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
MiniMax M3 Runtime Performance
Provider-specific output speed and catalog latency for MiniMax M3. Runtime does not affect the capability score.
Sort by
Provider | Output Speed | Catalog Latency | Max Input | Max Output | Updated |
|---|
| ProviderTogether | Output Speed213.561 tok/s | Catalog Latency0.962 s | Max Input1M | Max Output131.1K | Updated |
| ProviderMiniMax | Output Speed55.078 tok/s | Catalog Latency0.891 s | Max Input1M | Max Output1M | Updated |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
MiniMax M3 Specifications
Technical details for the model's default version.
- Version
- MiniMax M3
- Released
- Jun 1, 2026
- Knowledge cutoff
- Unknown
- Parameters
- 428B
- Context window
- 512K
- Max output
- 128K
- Inputs
- image, text, video
- Outputs
- text
- Open weights
- Yes
- License
- MIT
MiniMax M3 Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
Sort by
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionMiniMax M3 | Released | LLMBoard71.1 | Parameters428B | Context512K | Max output128K | Open weightsYes | LicenseMIT |
MiniMax M3 vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to MiniMax M3
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
MI
MiniMax M2.7
MiniMax
MI
MiniMax M2.5
MiniMax
MI
MiniMax M2.1
MiniMax
MI
MiniMax M2
MiniMax
MI
MiniMax M1 80K
MiniMax
MI
MiniMax M1 40K
MiniMax
What is MiniMax M3?
Key information about MiniMax M3 and its available data.
MiniMax M3 is the first open-weight model to combine three frontier capabilities: top-tier coding and agentic performance, a 1M-token context window, and native multimodality.
It is powered by MiniMax Sparse Attention (MSA), a new sparse attention architecture that partitions the KV cache into blocks to cut per-token compute at long context — roughly 1/20 the cost of the previous generation at 1M tokens, with more than 9x faster prefill and more than 15x faster decode while matching full attention on most capabilities.
Trained with mixed-modality data from step zero across 100T+ tokens, M3 natively supports image and video input and can operate a desktop computer. 7. M3 supports toggling thinking on or off at request time.
Data as of 2026-08-17.
FAQ
Common questions about MiniMax M3.
When was MiniMax M3 released?
MiniMax M3's default version was released on Jun 1, 2026.
How much does MiniMax M3 cost?
MiniMax M3's official API price is $0.30 per million input tokens and $1.2 per million output tokens via MiniMax (minimax.io). The lowest tracked third-party offer starts at $0.225 input and $0.90 output via EmpirioLabs AI.
Who created MiniMax M3?
MiniMax M3 was created by MiniMax.
What is the context window for MiniMax M3?
The default version has a 512K token context window.
Is MiniMax M3 open weight?
Yes. The default version is marked as open weight under MIT.
How many API providers offer MiniMax M3?
36 provider offerings are linked to the default version.
What models should I compare MiniMax M3 with?
Nearby ranked alternatives include Claude Sonnet 4.6, Sakana Namazu, GPT-5.2-Pro.