Skip to main content

Hugging Face

Prefer guided AI setup?

Instead of following these steps manually, you can use the AI Credential Assistant — describe what you want to connect and the AI will walk you through the setup step by step, tailored to your account.

Overview​

Connect your Hugging Face account to Nirvai to access thousands of pre-trained AI models for natural language processing, computer vision, audio, and more through the Inference API.

Supported Authentication Methods​

  • API Token - Required

Prerequisites​

Setup Instructions​

Step 1: Create a Hugging Face Account​

  1. Go to Hugging Face
  2. Sign up for a free account or log in

Step 2: Get Your API Token​

  1. Go to your profile Tokens page
  2. You'll see your existing token or can create a new one
  3. Copy the token (it should begin with hf_)

Step 3: Add to Nirvai​

  1. In Nirvai, go to create a new Hugging Face credential
  2. Paste your API token in the API Key field
  3. Save the credential

Features​

Hugging Face provides access to:

  • 350,000+ models for various tasks
  • Text generation and chat models
  • Embeddings for semantic search
  • Image generation and classification
  • Audio transcription and generation

Token Types​

  • Read tokens: Access public models and your private models
  • Write tokens: Upload models and datasets
  • Fine-grained tokens: Control access per repository

Troubleshooting​

Token Not Working​

  • Verify the token starts with hf_
  • Check if the token has the required permissions
  • Ensure the token hasn't been revoked

Model Access Denied​

  • Some models require accepting terms of use
  • Visit the model page and accept the license
  • Private models require appropriate token permissions

Rate Limit Errors​

  • Free tier has request limits
  • Consider upgrading to Pro for higher limits
  • Use the dedicated Inference Endpoints for production

Documentation adapted for Nirvai platform

Next Steps​

Got your credential? Connect it to an agent

Use the AI Tool Creator to turn this credential into agent-callable tools. Describe what your agent should be able to do and the AI investigates the API, configures the endpoints, and gives you a reviewable toolkit before saving anything.