// Open-model VRAM table · family
MiniMax: local deployment & VRAM
11 open-weight models from MiniMax (15 official repos incl. quantised builds). At 4-bit they range from 155 GB (MiniMax-M2.7) to 305 GB (MiniMax-VL-01).
official orgs: MiniMaxAIlatest 2026-08-07memory = estimate
01Text LLM (7)
| Model | Params | Context | BF16 | 4-bit | Smallest GPU (4-bit) | Mac (4-bit) | Licence | Released |
|---|---|---|---|---|---|---|---|---|
| MiniMax-M2.7 | 228.7B | 200K | 506 GB | 155 GB | 4 × RTX PRO 6000 (384GB) | Mac 256GB | Custom licence – read it | 2026-04-09 |
| MiniMax-M2.5 | 228.7B | 192K | 506 GB | 155 GB | 4 × RTX PRO 6000 (384GB) | Mac 256GB | Commercial with conditions | 2026-02-12 |
| MiniMax-M2.1 | 228.7B | 192K | 506 GB | 155 GB | 4 × RTX PRO 6000 (384GB) | Mac 256GB | Commercial with conditions | 2025-12-20 |
| MiniMax-M2 | 230B / 10B | 192K | 509 GB | 156 GB | 4 × RTX PRO 6000 (384GB) | Mac 256GB | Commercial with conditions | 2025-10-22 |
| MiniMax-M1-80k | 456B / 45.9B | — | 1.00 TB | 305 GB | 4 × RTX PRO 6000 (384GB) | Mac 512GB | Commercial OK | 2025-06-13 |
| MiniMax-M1-40k | 456B / 45.9B | — | 1.00 TB | 305 GB | 4 × RTX PRO 6000 (384GB) | Mac 512GB | Commercial OK | 2025-06-05 |
| MiniMax-Text-01-hf | 456B / 45.9B | 4M | 1.00 TB | 305 GB | 4 × RTX PRO 6000 (384GB) | Mac 512GB | Custom licence – read it | 2025-06-03 |
02Vision-language (2)
| Model | Params | Context | BF16 | 4-bit | Smallest GPU (4-bit) | Mac (4-bit) | Licence | Released |
|---|---|---|---|---|---|---|---|---|
| MiniMax-M3 | 428B / 23B | 1M | 942 GB | 286 GB | 4 × RTX PRO 6000 (384GB) | Mac 512GB | Custom licence – read it | 2026-06-02 |
| MiniMax-VL-01 | 456.4B | 8K | 1.00 TB | 305 GB | 4 × RTX PRO 6000 (384GB) | Mac 512GB | Licence not stated | 2025-01-12 |
03Text-to-speech / audio (1)
| Model | Params | BF16 | 4-bit | Smallest GPU (4-bit) | Mac (4-bit) | Licence | Released |
|---|---|---|---|---|---|---|---|
| MiniMax-Music3 | 2.43B | 5.9 GB | 2.1 GB | 8GB GPU | Mac 16GB | Licence not stated | 2026-08-07 |
04Video generation (1)
| Model | Params | BF16 | 4-bit | Smallest GPU (4-bit) | Mac (4-bit) | Licence | Released |
|---|---|---|---|---|---|---|---|
| MiniMax-H3 | 33.1B | 73.4 GB | 22.6 GB | RTX 5090 32GB | Mac 64GB | Custom licence – read it | 2026-07-28 |
05Other families
Qwen · IBM Granite · NVIDIA Nemotron · Gemma · Mistral / Magistral / Devstral · GLM / CogVideoX (Z.ai) · DeepSeek · Ai2 OLMo / Molmo · Ant Ling / Ring / Ming · Liquid LFM · MiniCPM / VoxCPM · Tencent Hunyuan / Hy · Phi · Llama · Meituan LongCat · Jina embeddings & rerankers · Wan · Kimi · Cohere Command / Aya · Xiaomi MiMo · FLUX · Microsoft VibeVoice / Fara / Harrier · InternLM / Intern-S · StepFun Step · Baidu ERNIE · NVIDIA Parakeet / Canary speech · Stable Diffusion / Stable Audio · Yi (01.AI) · ByteDance Seed · SmolLM / SmolVLM · BGE embeddings & rerankers · LTX Video · FunAudioLLM speech · gpt-oss · poolside Laguna · Whisper · Baichuan · Thinking Machines Inkling · xAI Grok · rednote dots
Totals include 10% + 0.5GB overhead and one 8K-token KV cache for text models. Estimates only. Method →