Index / providers / hugging-face

Hugging Face Inference Providers

Hugging Face's OpenAI-compatible router, forwarding requests to third-party hosts. One token reaches thousands of community abliterated, Heretic, Dolphin, Hermes, and roleplay models, almost all served by Featherless.

Models10
Per-token price$0.050–$0.70/1M tok
Plans from$9/mo
Free tierNo
huggingface.co ↗Pending review
[ 01 ] Models & prices

BILLINGBilled at each upstream provider's per-token rate with no markup. Routes served by Featherless don't show a per-token price.

ModelMethodContextIn $/1MOut $/1MPer 1K chats
Sao10K/L3-8B-Stheno-v3.2Llama 3 8B · 8B
Fine-tuned8K$0.050$0.050$0.075
NousResearch/Hermes-3-Llama-3.1-70BLlama 3.1 70B · 70B · Llama 3.1 Community
Fine-tuned131K$0.70$0.70$1.05
dphn/Dolphin-Mistral-24B-Venice-EditionMistral Small 24B · 24B · Apache-2.0
Fine-tuned—Included in plan
huihui-ai/Llama-3.3-70B-Instruct-abliteratedLlama 3.3 70B · 70B · Llama 3.3 Community
Abliterated—Included in plan
huihui-ai/Huihui-Qwen3.5-27B-abliteratedQwen3.5 27B · 27B · Apache-2.0
Abliterated—Included in plan
Abliterated—Included in plan
Abliterated—Included in plan
ArliAI/Qwen3.5-27B-DerestrictedQwen3.5 27B · 27B · Apache-2.0
Abliterated—Included in plan
Sao10K/L3.3-70B-Euryale-v2.3Llama 3.3 70B · 70B · Llama 3.3 Community
Fine-tuned—Included in plan
Fine-tuned—Included in plan

Per 1K chats = cost of 1,000 messages of ~1,000 tokens in and ~500 out. Click a model to compare it across providers.All models →

[ 02 ] At a glance
Method
Mixed
Access
Web chat + OpenAI-compatible API
Logging
No logs
Trains on prompts
No
Guarantee
Policy only
Jurisdiction
US
Signup
Account required
Payments
Card
Data checked
Oct 11, 2026

PRIVACYSays it doesn't store request or response bodies when routing (debug logs up to 30 days without user data). Each upstream provider's policy also applies. Policy ↗

[ 03 ] Scores
Compliance
Benchmark pending
measured
Capability
Benchmark pending
measured
Privacy
58.0
from published policy & pricing
Truthfulness
Benchmark pending
measured
Cost
100.0
from published policy & pricing
Speed
Benchmark pending
measured

Composite: pending capability and compliance runs · method

[ 04 ] Notes & sources

Mostly an access route to Featherless, Novita, and DeepInfra (all listed separately). HuggingChat is still live and includes Stheno v3.2 and Hermes 3 70B.

  1. https://router.huggingface.co/v1/models
  2. https://huggingface.co/docs/inference-providers/pricing
  3. https://huggingface.co/docs/inference-providers/security
  4. https://huggingface.co/chat/api/v2/models