Local Models by VRAM: Quick Reference
Open-weight models covered by Local Model Watch, grouped by the smallest amount of GPU memory (VRAM) they are estimated to run in. Each figure is computed by this site from the actual size of the distributed files plus a runtime overhead — not quoted from model cards. See our editorial policy for the method.
Models listed under a tier do not fit in the smaller tiers. If you have a 24GB GPU, the 8GB, 12GB, 16GB and 24GB tables all apply to you. Models that don’t fit even the largest tier (80GB) are listed separately at the bottom, since that too is something readers need to know.
The Runs on column lists engines (Ollama / LM Studio via llama.cpp, vLLM, MLX) confirmed to run the model today, matched from each engine’s own model registry and the GGUF/MLX builds we have found; it does not apply to image, video and audio models. “-" means either not applicable or not yet confirmed — see the model’s own article for the reasoning.
Last updated 2026-11-02 (JST). 52 models listed.
Everything that fits your GPU: 8GB / 12GB / 16GB / 24GB
4GB (laptop iGPU / phone class)
→ Scroll horizontally to see all columns
| Model | Parameters | Smallest build | Est. memory | Runs on | Article |
|---|---|---|---|---|---|
| LiquidAI/LFM2.5-350M-Diffusion-Exp | 425M | BF16 (0.8GB) | 0.9GB | – | Read |
| Cloudflare/clef-flash | 9.4B | IQ2_M (3.3GB) | 4.0GB | Ollama / LM Studio, vLLM, MLX | Read |
| FermionResearch/Phonon-2 | 627M | Q8_0 (0.7GB) | 0.8GB | – | Read |
| apple/LensVLM-9B | 9.4B | IQ2_M (3.3GB) | 4.0GB | Ollama / LM Studio, vLLM, MLX | Read |
| togethercomputer/Tev1-0.8B-experimental | 873M | F32 (1.6GB) | 2.0GB | vLLM | Read |
| LiquidAI/LFM2.5-VL-3B-DSpark | 279M | F16 (0.5GB) | 0.6GB | – | Read |
| togethercomputer/Tev1-4B-experimental | 4.7B | Q6_K_S (3.4GB) | 4.0GB | Ollama / LM Studio, vLLM | Read |
| pfnet/plamo-3-610m-fin-instruct | 890M | F16 (1.7GB) | 2.0GB | Ollama / LM Studio, vLLM | Read |
| Cactus-Compute/needle3 | – | Original precision (0.2GB) | 0.3GB | – | Read |
| tencent/WeVisDoc-4B | 4.4B | Q5_K_S (2.9GB) | 3.5GB | – | Read |
| harshatheg/Qwen-2.5-1B-RLCD | 1.5B | BF16 (2.9GB) | 3.5GB | MLX | Read |
| Efficient-Large-Model/H3-to-LTX-Latent-Adapter | 195M | BF16 (0.4GB) | 0.4GB | – | Read |
| openbmb/MiniCPM5-2B | 2.5B | Q8_0 (2.5GB) | 3.0GB | Ollama / LM Studio, vLLM, MLX | Read |
| phasefield-audio/Irodori-TTS-v4.1-Anime | 766M | F32 (2.9GB) | 3.4GB | – | Read |
8GB (RTX 4060 / 3060 Ti, etc.)
→ Scroll horizontally to see all columns
| Model | Parameters | Smallest build | Est. memory | Runs on | Article |
|---|---|---|---|---|---|
| google/DiarizationLM-Gemma-4-E4B-v1 | 8.0B | Q4_K_M (4.9GB) | 5.9GB | Ollama / LM Studio, vLLM | Read |
| Qwen/Qwen3.8-27B | 27.8B | IQ1_M (6.3GB) | 7.5GB | Ollama / LM Studio, vLLM, MLX | Read |
| prism-ml/Ternary-Bonsai-2-27B-gguf | 27.8B | Q1_0 (5.5GB) | 6.6GB | Ollama / LM Studio, MLX | Read |
12GB (RTX 4070 / 3060 12GB, etc.)
→ Scroll horizontally to see all columns
| Model | Parameters | Smallest build | Est. memory | Runs on | Article |
|---|---|---|---|---|---|
| Cloudflare/clef | 27.4B | Q2_K (9.9GB) | 11.8GB | Ollama / LM Studio, vLLM, MLX | Read |
| Altworld/Hemmingway-1 | 26.9B | IQ2_M (9.8GB) | 11.8GB | Ollama / LM Studio, vLLM, MLX | Read |
| Edge0/Audio8-ASR-Infinite | 4.1B | BF16 (7.6GB) | 9.1GB | – | Read |
| ggml-org/MiMo-V2.6-Distill-Qwen-9B-GGUF | 9.4B | Q8_0 (8.9GB) | 10.6GB | Ollama / LM Studio | Read |
| bartowski/vectionlabs_Salience-27B-R6-GGUF | 27.8B | IQ2_M (9.8GB) | 11.8GB | Ollama / LM Studio | Read |
| tencent/Simple-Attention-Sparsification | 4.0B | BF16 (7.5GB) | 9.0GB | – | Read |
| bartowski/TheDrummer_Orion-26B-A4B-v1.1-GGUF | 25.8B | IQ2_S (9.6GB) | 11.5GB | Ollama / LM Studio | Read |
| Comfy-Org/YuE2 | 3.6B | BF16 (6.8GB) | 8.1GB | – | Read |
| bartowski/Gryphe_Pantheon-Reasoning-26B-A4B-1.1-V2-GGUF | 26.5B | IQ2_S (10.0GB) | 12.0GB | Ollama / LM Studio | Read |
| m-a-p/YuE2-3B | 3.6B | BF16 (6.8GB) | 8.1GB | – | Read |
16GB (RTX 5060 Ti 16GB / 4060 Ti 16GB, etc.)
→ Scroll horizontally to see all columns
| Model | Parameters | Smallest build | Est. memory | Runs on | Article |
|---|---|---|---|---|---|
| orcarouter/OrcaSAQ-2-27B | 27.8B | EXL3 (11.4GB) | 13.7GB | vLLM | Read |
| agentionai/Signal-3.8-27B-GGUF | 27.8B | IQ4_XS (13.3GB) | 15.9GB | Ollama / LM Studio | Read |
| nex-agi/Nex-N2.5-mini | 35.1B | Q2_K (12.9GB) | 15.5GB | Ollama / LM Studio, vLLM, MLX | Read |
| Jackrong/Qwopus3.8-27B-Flash-GGUF | 27.8B | Q3_K_M (12.6GB) | 15.1GB | Ollama / LM Studio | Read |
24GB (RTX 4090 / 3090, etc.)
→ Scroll horizontally to see all columns
| Model | Parameters | Smallest build | Est. memory | Runs on | Article |
|---|---|---|---|---|---|
| nvidia/PixelUMM | 8.2B | BF16 (15.3GB) | 18.3GB | – | Read |
| ggml-org/OpenJev-GGUF | 27.4B | Q4_K_M (17.7GB) | 21.2GB | Ollama / LM Studio | Read |
| XingChen-AGI/Xing4.0-29B-A4B | 31.2B | IQ4_NL (18.7GB) | 22.5GB | Ollama / LM Studio, MLX | Read |
| inclusionAI/Realtime-Venus | – | Original precision(Realtime-Venus-Audio) (17.5GB) | 20.9GB | – | Read |
| EleutherAI/olmo3-7b-sdf-sft-clean150 | 7.3B | BF16 (13.6GB) | 16.3GB | vLLM | Read |
| Edge0/Edge0-35B-A3B-preview | 36.0B | MLX 4bit (18.2GB) | 21.8GB | vLLM, MLX | Read |
32GB (RTX 5090, etc.)
→ Scroll horizontally to see all columns
| Model | Parameters | Smallest build | Est. memory | Runs on | Article |
|---|---|---|---|---|---|
| MiniMaxAI/MiniMax-H3 | – | Q3_K_XL (26.5GB) | 31.8GB | – | Read |
| IFM/K2-Horizon-32B-NVFP4 | – | NVFP4 (21.7GB) | 26.0GB | vLLM | Read |
| IFM/K2-Horizon-MoVA-36B-A4B-GGUF | 37.4B | Q5_K_M (24.6GB) | 29.5GB | Ollama / LM Studio | Read |
80GB class (A100 / H100)
→ Scroll horizontally to see all columns
| Model | Parameters | Smallest build | Est. memory | Runs on | Article |
|---|---|---|---|---|---|
| elyza/ELYZA-Thinking-1.0-llm-jp-4-32b-a3b | 32.1B | BF16 (59.9GB) | 71.8GB | vLLM | Read |
Over 80GB (datacenter GPU / multi-GPU required)
→ Scroll horizontally to see all columns
| Model | Parameters | Smallest build | Est. memory | Runs on | Article |
|---|---|---|---|---|---|
| Qwen/Qwen3.8-Flash-Next | 180.0B | BF16 (335.3GB) | 402.3GB | vLLM | Read |
| NaiveAI/Naive-N0.5-Flash | – | Original precision (575.4GB) | 690.5GB | – | Read |
| ggml-org/GLM-5.3-Flash-GGUF | 321.3B | Q2_K (124.6GB) | 149.5GB | Ollama / LM Studio | Read |
| XiaomiMiMo/MiMo-V2.6-Pro-RL | – | Q2_K (347.8GB) | 417.4GB | Ollama / LM Studio, vLLM | Read |
| deepseek-ai/DeepSeek-V4-Pro-0813 | 1650.5B | FP8 (831.4GB) | 997.7GB | Ollama / LM Studio, vLLM | Read |
| IFM/K2-Horizon-375B-A23B-NVFP4 | – | NVFP4 (213.8GB) | 256.6GB | vLLM | Read |
| ggml-org/MiMo-V2.6-Flash-RL-GGUF | – | Q2_K (117.5GB) | 141.1GB | Ollama / LM Studio | Read |
| bartowski/Intern-S2-397B-GGUF | 403.4B | IQ1_S (85.3GB) | 102.4GB | Ollama / LM Studio | Read |
| deepseek-ai/DeepSeek-V4.1-Flash | 763.2B | FP8 (475.3GB) | 570.3GB | vLLM | Read |
| nvidia/DeepSeek-V4-Pro-0813-nvfp4-DSpark | 1650.5B | NVFP4 (837.4GB) | 1004.9GB | vLLM | Read |
| nex-agi/Nex-N2.5-Pro | 396.8B | IQ1_S (81.8GB) | 98.2GB | Ollama / LM Studio, vLLM | Read |