Check my GPU

LLMs ·

SmolVLM 500M Instruct

Parameters
0.5B
Architecture
Dense
Max context
8K tokens
License
apache-2.0
Released
2025-01

How to run it

Quantized files · GGUF

Works in LM Studio, llama.cpp, Jan and Ollama. No sign-in needed.

huggingface.co/ggml-org/SmolVLM-500M-Instruct-GGUF (opens in a new tab)

ollama run hf.co/ggml-org/SmolVLM-500M-Instruct-GGUF

Original weights · Hugging Face

Full-precision files from Hugging Face, for fine-tuning, vLLM, Transformers.

huggingface.co/HuggingFaceTB/SmolVLM-500M-Instruct (opens in a new tab)

No sign-in needed.

Links checked 5 days ago against Hugging Face, Ollama and GitHub.

Memory needed

Weights plus KV cache and runtime overhead. Leave a little extra free for the operating system and other apps.

Quantization Quality Weights Total at 8K context Download
Q2_K Low 0.2 GB 1.1 GB GGUF files (opens in a new tab)
Q3_K_M Low 0.2 GB 1.2 GB GGUF files (opens in a new tab)
Q4_K_M Good 0.3 GB 1.2 GB GGUF files (opens in a new tab)
Q5_K_M High 0.4 GB 1.3 GB GGUF files (opens in a new tab)
Q6_K High 0.4 GB 1.4 GB GGUF files (opens in a new tab)
Q8_0 Very high 0.5 GB 1.5 GB GGUF (opens in a new tab)
F16 Full 1.0 GB 2.0 GB GGUF (opens in a new tab) Original weights (opens in a new tab)

Best hardware for SmolVLM 500M Instruct

The least expensive GPU in our catalog that runs it without offloading is the Intel Arc A380 (launch price $139) at F16.

Grade · points Hardware Quant Memory Speed Fit
51 ptsUsable (C) · 51 of 100 points: Model quality 4.5/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 6.8/15 NVIDIA GeForce RTX 4060 F16 2.4 GB / 8.0 GB 154 tok/s Runs well
50 ptsUsable (C) · 50 of 100 points: Model quality 4.5/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 6.1/15 Apple M4 (16 GB) F16 2.8 GB / 10.7 GB 58.4 tok/s Runs well
49 ptsUsable (C) · 49 of 100 points: Model quality 4.5/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 5.0/15 Apple M1 (16 GB) Q8_0 2.3 GB / 10.7 GB 72.0 tok/s Runs well
49 ptsUsable (C) · 49 of 100 points: Model quality 4.5/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 4.9/15 Intel Arc B580 F16 2.6 GB / 12.0 GB 197 tok/s Runs well
49 ptsUsable (C) · 49 of 100 points: Model quality 4.5/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 4.9/15 NVIDIA GeForce RTX 3060 12GB F16 2.6 GB / 12.0 GB 203 tok/s Runs well
49 ptsUsable (C) · 49 of 100 points: Model quality 4.5/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 4.9/15 NVIDIA GeForce RTX 4070 SUPER F16 2.6 GB / 12.0 GB 285 tok/s Runs well
49 ptsUsable (C) · 49 of 100 points: Model quality 4.5/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 4.7/15 Apple M4 (24 GB) F16 3.2 GB / 16.1 GB 58.4 tok/s Runs well
49 ptsUsable (C) · 49 of 100 points: Model quality 4.5/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 4.7/15 Apple M4 Pro (24 GB) F16 3.2 GB / 16.1 GB 133 tok/s Runs well
48 ptsUsable (C) · 48 of 100 points: Model quality 4.5/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 4.0/15 AMD Radeon RX 9070 XT F16 2.8 GB / 16.0 GB 311 tok/s Runs well
48 ptsUsable (C) · 48 of 100 points: Model quality 4.5/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 4.0/15 NVIDIA GeForce RTX 4060 Ti 16GB F16 2.8 GB / 16.0 GB 163 tok/s Runs well

Check it on your own hardware

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.