From the model card
Fields below are copied from the tags and counters on the HuggingFace repository openai/gpt-oss-120b at our last fetch. They are set by the uploader, not verified by us; rows with no tag are omitted. How this page is made.
- Publisher (HF namespace)
- openai
- Pipeline tag
- text-generation
- Library
- Transformers, vLLM
- Weight formats
- safetensors
- License tag
apache-2.0— read the license file in the repo before relying on it- Papers cited
- arXiv:2508.10925
- Downloads (HF counter at last fetch)
- 5,174,914
- Likes (HF counter at last fetch)
- 5,148
- Model card
- https://huggingface.co/openai/gpt-oss-120b
Use cases
- Self-hosted chat assistants requiring large-model quality
- Batch document processing on GPU clusters
- Fine-tuning base for domain-specific applications
- Research comparing open versus proprietary model behavior
Pros
- Apache 2.0 license allows unrestricted commercial use
- MXFP4 support reduces VRAM requirements at inference scale
- vLLM compatible for high-throughput production serving
Cons
- 120B scale requires 4–8 high-VRAM GPUs for full-precision inference
- Text-only — no multimodal capability
- Community fine-tunes and GGUF quants lag behind smaller popular models
Tags
transformerssafetensorsgpt_osstext-generationvllmconversationalarxiv:2508.10925license:apache-2.0eval-resultsendpoints_compatible8-bitmxfp4deploy:sagemakerdeploy:azureregion:us