SmolLM GGUF models
1353 GGUF models in the SmolLM family, quantized by 429 uploaders. Sizes run from 2M to 20.4B parameters. The longest trained context is 128K tokens.
About the SmolLM GGUF models
SmolLM is represented on ggufmodels.com by 1,353 GGUF models, quantized by 429 uploaders on Hugging Face. The most downloaded are TwIL LM3, SmolLM2 135M Instruct and SmolLM2 135M Instruct. Parameter counts run from 2M to 20.4B, and the longest trained context window is 128K tokens.
A Q4_K_M file of a SmolLM model is between 0.0 to 14.1 GB, which is the quant most people start with. Most files come from RichardErkhov (280), mradermacher (248) and tensorblock (19). The quants you will meet most often are Q4_K_M (961), Q8_0 (863), Q5_K_M (707) and Q6_K (671), where the number is how many models offer that quant.
Frequently asked questions
Which SmolLM GGUF should I download?
The most downloaded SmolLM GGUF here is TwIL LM3. Pick the size that fits your GPU first, then choose a quant: Q4_K_M is the usual starting point, and Q5_K_M or Q6_K give more quality if you have memory to spare. The VRAM pages below list what fits each GPU.
How much VRAM does a SmolLM model need?
Q4_K_M files of SmolLM models on this site range from 0.0 to 14.1 GB. Add the KV cache for your context length and about 0.5 GB for the runtime, which usually means 1.5 to 3 GB on top of the file at an 8K context.
Where do SmolLM GGUF files come from?
The most common quantizers here are RichardErkhov, mradermacher and tensorblock. They convert the original model and publish quantized files on Hugging Face. ggufmodels.com does not host files; every download button opens Hugging Face.