Pricing
Plans for community and organizations
👩🔬 Community
- Contributor
- Free
-
Upload public models
-
Community support
-
Follow tags for new model alerts
-
Soon: compare models with model hub history
- Become a 🤗 Supporter
- $9/mo
-
Get a PRO badge on the website and the forum
Show your support for the best community in ML 😃
-
Upload up to 5 private models
-
Community support
-
Early preview access to all the upcoming hub features
-
Soon: compare models with model hub history
🚌 Organizations
- Lab
- $199/mo
-
Upload and serve private models
-
Create and serve AutoNLP models
Training costs billed separately
-
Accelerated Inference API: CPU
2x faster than inference widgets
-
10M inference input characters /mo
Additional characters $15/M
-
1 pinned model
Instantly available for inference
-
Email support
- Startup
- $599/mo
-
Upload and serve private models
-
Create and serve AutoNLP models
Training costs billed separately
-
Accelerated Inference API: CPU + GPU
2x-10x faster than inference widgets
-
50M inference input characters /mo
Additional characters $10/M
-
5 pinned models
Instantly available for inference
-
Email support
- Enterprise
- Custom quote
-
We will design a custom solution to your requirements
For volumes of 100M+ characters/mo, contact us for volume-based tiered pricing.
If your company is building its own infrastructure, our Expert Acceleration Program can support your team, and our Containerized Solutions can be deployed on-premise.
-
Priority support and access to the Hugging Face team
Private models and datasets
Upload models and datasets privately to your account or
organization, access them with Transformers,
Datasets
and Tokenizers, and manage them in the hub with
built-in version control using git and
git-lfs.
AutoNLP
Create, evaluate and deploy state of the art models fine-tuned to your training data in three simple steps: select a task, a language, upload your data, and let AutoNLP do its magic.
Accelerated Inference API
Ship NLP features faster using our hosted infrastructure, built for production scale with 2x-10x accelerated inference on 7,000+ state of the art models, or your private models.