// Open-model VRAM table · family
ByteDance Seed: local deployment & VRAM
9 open-weight models from ByteDance Seed (13 official repos incl. quantised builds). At 4-bit they range from 6.5 GB (UI-TARS-1.5-7B) to 26.8 GB (Seed-OSS-36B-Instruct).
official orgs: ByteDance-Seedlatest 2025-08-20memory = estimate
01Text LLM (2)
| Model | Params | Context | BF16 | 4-bit | Smallest GPU (4-bit) | Mac (4-bit) | Licence | Released |
|---|---|---|---|---|---|---|---|---|
| Seed-OSS-36B-Instruct | 36.2B | 512K | 82.2 GB | 26.8 GB | RTX 5090 32GB | Mac 64GB | Commercial OK | 2025-08-20 |
| Seed-Coder-8B-Instruct | 8.25B | 32K | 19.7 GB | 7.1 GB | 8GB GPU | Mac 16GB | Commercial OK | 2025-04-27 |
02Vision-language (1)
| Model | Params | Context | BF16 | 4-bit | Smallest GPU (4-bit) | Mac (4-bit) | Licence | Released |
|---|---|---|---|---|---|---|---|---|
| UI-TARS-1.5-7B | 8.29B | 125K | 19.2 GB | 6.5 GB | 8GB GPU | Mac 16GB | Commercial OK | 2025-04-16 |
03Omni / audio LLM (1)
| Model | Params | Context | BF16 | 4-bit | Smallest GPU (4-bit) | Mac (4-bit) | Licence | Released |
|---|---|---|---|---|---|---|---|---|
| BAGEL-7B-MoT | 14.7B | 32K | 33.3 GB | 10.8 GB | 12GB GPU | Mac 24GB | Commercial OK | 2025-05-19 |
04Video generation (3)
| Model | Params | BF16 | 4-bit | Smallest GPU (4-bit) | Mac (4-bit) | Licence | Released |
|---|---|---|---|---|---|---|---|
| SeedVR-3B ↗ | 3B | 7.1 GB | 2.5 GB | 8GB GPU | Mac 16GB | Commercial OK | 2025-06-11 |
| SeedVR2-7B | 7B | 15.9 GB | 5.2 GB | 8GB GPU | Mac 16GB | Commercial OK | 2025-06-11 |
| SeedVR2-3B | 3B | 7.1 GB | 2.5 GB | 8GB GPU | Mac 16GB | Commercial OK | 2025-06-10 |
05Translation (2)
| Model | Params | Context | BF16 | 4-bit | Smallest GPU (4-bit) | Mac (4-bit) | Licence | Released |
|---|---|---|---|---|---|---|---|---|
| Seed-X-Instruct-7B ↗ | 7.51B | 32K | 18.1 GB | 6.6 GB | 8GB GPU | Mac 16GB | Commercial OK | 2025-07-16 |
| Seed-X-PPO-7B ↗ | 7.51B | 32K | 18.1 GB | 6.6 GB | 8GB GPU | Mac 16GB | Commercial OK | 2025-07-16 |
06Other 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 · MiniMax · Baidu ERNIE · NVIDIA Parakeet / Canary speech · Stable Diffusion / Stable Audio · Yi (01.AI) · 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 →