llm-jp models
2 llm-jp models from llm-jp on Hugging Face, from 152M to 33.2B parameters, published by llm-jp in BF16, with contexts from 4K to 64K tokens. The smallest official model, llm-jp-3-150m, needs about 0.3 GB of VRAM at its published precision; the cheapest live fit is RTX 3070 at $0.088/hr.
Pick a size
One row per official llm-jp 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 |
|---|---|---|---|---|---|---|---|
| llm-jp-3-150m | 152M | 0.3 GB | 0.2 GB | 0.1 GB | RTX 3070 | $0.088/hr | 0.75 GB |
| llm-jp-4-33b-thinking | 33.2B | 74.3 GB | 37.1 GB | 18.6 GB | A100 | $1.18/hr | 8.0 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 (2)
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| llm-jp-3-150m | 152M | BF16 | 0.3 GB | RTX 3070 | $0.088/hr |
| llm-jp-4-33b-thinking | 33.2B | BF16 | 74.3 GB | A100 | $1.18/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.