HuggingFace Inference
online
Run 100,000+ open models. $0.10/month serverless free.
// At a glance
Free Tier
$0.10/month serverless · no card
// Free tier details
Available Models
100000+ community models
Monthly Requests
$0.10/month serverless credit
// Quick start
"text">-purple-400">curl https://api"text-amber-400">-inference.huggingface.co/models/mistralai/Mistral-7B"text-amber-400">-Instruct"text-amber-400">-v0.3 \
"text-amber-400">-H "Authorization: Bearer YOUR_HF_TOKEN" \
"text-amber-400">-H "Content">-Type: application/json" \
"text-amber-400">-d '{"inputs": "Hello, how are you?"}'
// Overview
Serverless inference for any model on the Hub. Includes free serverless CPU tier with $0.10/month credit and access to most popular models.
// Pros
- Largest model catalog in the world
- Includes embeddings, vision, audio
- Truly pay-as-you-go
// Cons
- Serverless cold starts
- Rate limits on free tier
// Score breakdown
Reliability (35%) (from 2m ago health check)100/100
Free Tier Generosity (30%) (computed from quota, no-CC, no-phone fields)85/100
Documentation (20%) (human rating)88/100
Popularity (15%) (GitHub stars (log-normalised), or manual baseline)90/100
Methodology: apivault.dev/methodology
// Best for
Niche modelsEmbeddingsOpen-source workflows
// Recent changes
May 30, 2026Faster cold starts on PRO planupdated