gpt-oss models
6 gpt-oss models from OpenAI on Hugging Face, from 350M to 116.8B parameters, published by OpenAI in MXFP4, with a 128K-token context. The smallest official model, gpt-oss-20b, needs about 16.0 GB of VRAM at its published precision; the cheapest live fit is RTX A4000 at $0.167/hr.
Pick a size
One row per official gpt-oss 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 |
|---|---|---|---|---|---|---|---|
| gpt-oss-20b | 20.9B | 16.0 GB | – | – | RTX A4000 | $0.167/hr | 0.75 GB |
| gpt-oss-120b | 116.8B | 80.0 GB | – | – | A100 | $1.31/hr | 1.1 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 |
|---|---|---|---|---|---|
| gpt-oss-20b | 20.9B | MXFP4 | 16.0 GB | RTX A4000 | $0.167/hr |
| gpt-oss-120b | 116.8B | MXFP4 | 80.0 GB | A100 | $1.31/hr |
Fine-tunes and community models (4)
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| EAGLE3-gpt-oss-20b | 350M | BF16 | 0.8 GB | RTX 3070 | $0.088/hr |
| gpt-oss-120b-Eagle3-short-context | 787M | BF16 | 1.8 GB | RTX 3070 | $0.088/hr |
| Huihui-gpt-oss-20b-BF16-abliterated | 20.9B | BF16 | 46.7 GB | RTX A6000 | $0.363/hr |
| gpt-oss-20b-BF16 | 20.9B | BF16 | 46.7 GB | RTX A6000 | $0.363/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.
More from OpenAI
- Whisper (30 models)