SmolLM-1 models
5 SmolLM-1 models from Hugging Face on Hugging Face, from 135M to 1.7B parameters, published by Hugging Face in F32 and BF16, with a 2K-token context. The smallest official model, SmolLM-135M-Instruct, needs about 0.3 GB of VRAM at its published precision; the cheapest live fit is RTX 4070 Super at $0.121/hr.
Part of the SmolLM series · Next generation: SmolLM2
Pick a size
One row per official SmolLM-1 size: the VRAM it needs at each precision, the cheapest GPU that holds it today, and the KV cache for a 32K-token context.
| Model | Parameters | Native VRAM | FP8 VRAM | INT4 VRAM | Live GPU fit (native) | Est. $/hr | KV cache at 32K |
|---|---|---|---|---|---|---|---|
| SmolLM-135M | 135M | 0.6 GB | 0.2 GB | 0.1 GB | V100 | $0.088/hr | 0.70 GB |
| SmolLM-360M-Instruct | 362M | 0.8 GB | 0.4 GB | 0.2 GB | RTX 4070 Super | $0.121/hr | 1.3 GB |
| SmolLM-1.7B | 1.7B | 7.7 GB | 1.9 GB | 1.0 GB | V100 | $0.088/hr | 6.0 GB |
VRAM is weights times a flat 1.2 overhead; the KV cache is a separate per-model figure at 16-bit, shown where the architecture is published. See the methodology.
Official models (4)
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| SmolLM-135M | 135M | F32 | 0.6 GB | V100 | $0.088/hr |
| SmolLM-135M-Instruct | 135M | BF16 | 0.3 GB | RTX 4070 Super | $0.121/hr |
| SmolLM-360M-Instruct | 362M | BF16 | 0.8 GB | RTX 4070 Super | $0.121/hr |
| SmolLM-1.7B | 1.7B | F32 | 7.7 GB | V100 | $0.088/hr |
Fine-tunes and community models (1)
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| SmolLM-135M-Instruct-FP32 | 135M | F32 | 0.6 GB | V100 | $0.088/hr |
VRAM is for the precision the model is published in: the weights times a flat 1.2 overhead, with the KV cache not included. See the methodology. The fit shown is the lowest-priced single GPU type that holds the model at that precision, or the lowest-priced multi-GPU set (up to 8) when none does.
