Spaces:

---

huggingface

/

HuggingDiscussions

like 91

Running

Love that I can call DeepSeek R1 directly from the Hub 🔥

from huggingface_hub import InferenceClient

client = InferenceClient(
    provider="together",
    api_key="xxxxxxxxxxxxxxxxxxxxxxxx"
)

messages = [
    {
        "role": "user",
        "content": "What is the capital of France?"
    }
]

completion = client.chat.completions.create(
    model="deepseek-ai/DeepSeek-R1", 
    messages=messages, 
    max_tokens=500
)

print(completion.choices[0].message)

Is it possible to set a monthly payment budget or rate limits for all the external providers? I don't see such options in billings tab. In case a key is or session token is stolen, it can be quite dangerous to my thin wallet:(

@benhaotang you already get spending notifications when crossing important thresholds ($10, $100, $1,000) but we'll add spending limits in the future

@benhaotang you already get spending notifications when crossing important thresholds ($10, $100, $1,000) but we'll add spending limits in the future

Thanks for your quick reply, good to know!

Would be great if you could add Nebius AI Studio to the list :) New inference provider on the market, with the absolute cheapest prices and the highest rate limits...

Could be good to add featherless.ai

TitanML !!

This is a great idea for Wan model inference. By the way, if you ever need to share HTML previews of your AI video outputs, the HTML to URL Converter is super handy for quick sharing without any signup.

Hi @julien-c @Wauplin @SBrandeis @hanouticelina,

We're the team behind Phala and we'd like to formally express our interest in joining the Hugging Face Inference Providers program.

About Phala

Phala (phala.com) is a confidential AI cloud that delivers private LLM inference on hardware-protected GPU infrastructure. Phala runs LLMs on dedicated GPU clusters with hardware-level isolation, providing runtime attestation so users can cryptographically verify that their prompts and outputs were never exposed. Our OpenAI-compatible API gateway at https://api.redpill.ai/v1 gives developers a drop-in replacement for the OpenAI SDK with verifiable privacy guarantees.

We are already a verified inference provider on OpenRouter (openrouter.ai/provider/phala), where we have processed over 2.9 billion tokens across 14 models as of April 2026 — demonstrating production-grade reliability and scale.

Why We're a Strong Fit

Unique Differentiation — Confidential Inference: Phala is the only inference provider in the current HF lineup that offers hardware-attested, confidential GPU inference. For users handling sensitive data, regulated workloads, or privacy-critical applications, this is a meaningful capability gap we can fill.

OpenAI-Compatible API: Our API (/v1/chat/completions, /v1/embeddings) is a full drop-in replacement for the OpenAI SDK — making JS and Python client integration straightforward.

Proven Scale on OpenRouter: Our top models by token volume on OpenRouter:

Supported Models & Tasks

We currently serve the following open-weight models on Phala's confidential GPU infrastructure:

Supported Tasks: Chat Completion (LLM), Chat Completion (VLM), Feature Extraction (Embeddings)

Integration Readiness

• Our API is fully OpenAI-compatible, so the JS and Python client integration should be straightforward.
• We have an existing organization on the Hub (huggingface.co/phalanetwork) and are ready to upgrade to a Team/Enterprise plan as required.
• We're prepared to submit the JS client PR, register model mappings via the Model Mapping API, implement the billing endpoint, and follow the full onboarding checklist.
• We can have a working integration ready within 1–2 weeks of receiving guidance.

Contact

• Website: https://phala.com
• API Base URL: https://api.redpill.ai/v1
• OpenRouter Profile: https://openrouter.ai/provider/phala
• Hub Org: https://huggingface.co/phalanetwork

We're happy to provide API access for testing, jump on a call, or align our implementation to any specific requirements. Looking forward to hearing from you!

Best,
The Phala Team

This is a really cool move for UomiRouter! Network reliability is so important for inference. By the way, if you ever need to check your connection stability, you can test packet loss online at packetlosstest.dev. Super handy for ensuring smooth performance.

New Inference Provider — GonkaBroker (gonkabroker)

Hi! We'd like to be listed as an Inference Provider. We route to the Gonka.ai decentralized inference network via an OpenAI-compatible API.

• Org / provider slug: gonkabroker (Team plan active)
• JS client PR: https://github.com/huggingface/huggingface.js/pull/2226 (conversational + text-generation)
• Base URL: https://proxy.gonkabroker.com (/v1/chat/completions, /v1/completions, /v1/models)
• /v1/models exposes per-model pricing (USD/M) + context length for the comparison table / :cheapest.
• Billing API: POST https://app.gonkabroker.com/api/partners/huggingface/billing, Bearer auth, {"requestIds":[...]} → {"requests":[{"requestId","costNanoUsd"}]} (nano-USD).
• Request ID header: Inference-Id on every response.
• Logo (light + dark) ready to send.

Could you enable the org server-side so we can register model mappings? Happy to share the master key for routed traffic over a private channel. Thanks!

Following up on the GonkaBroker provider request above — the JS PR is open and ready for review: https://github.com/huggingface/huggingface.js/pull/2226 (CI awaiting maintainer approval to run). Paid Team plan active, all prerequisites done. Would love help getting it moving.

New Inference Provider integration — deAPI

Hi HF team 👋

We're deAPI (https://deapi.ai, HF org: https://huggingface.co/deAPI-ai) and we'd like to register as an official Inference Provider on the Hub.

deAPI provides an OpenAI-compatible inference API at:

https://oai.deapi.ai/v1

with Bearer token authentication. The endpoint works with the official OpenAI SDK by changing the base_url, and we also maintain a native Python SDK for deAPI workflows.

For the initial integration, we'd like to start with image, audio and embedding tasks:

• text-to-image

  • FLUX.2 Klein 4B BF16
  • Z-Anime
  • Z-Image-Turbo INT8
  • FLUX.1 schnell

• image-to-image

  • FLUX.2 Klein 4B BF16
  • Qwen Image Edit Plus

• automatic-speech-recognition

  • Whisper Large V3

• text-to-speech

  • Qwen3 TTS CustomVoice
  • Qwen3 TTS VoiceDesign
  • Qwen3 TTS VoiceClone
  • Chatterbox
  • Kokoro

• feature-extraction / embeddings

  • BGE M3

Exact model slugs, limits and defaults can be resolved through our Model Selection endpoint during the mapping setup.

LLM/chat models are also on our near-term roadmap, and we expect to add conversational / text-generation support shortly after the initial provider setup is live.

We've read the “register as a provider” guide and have already opened the following huggingface.js PR:

https://github.com/huggingface/huggingface.js/pull/2240

The PR starts with text-to-image on a single model in staging. We're also adding a per-response correlation header using Inference-Id with a UUID value, so it can support routed billing and request reconciliation.

One clarification on plan requirements before we finalize the setup. We understand a Team/Enterprise org plan is needed for the Model Mapping API (the routed path). Is the same plan required for a custom-key-only listing (clientSideRoutingOnly, like OpenAI) — where users add their own deAPI key and we bill them directly? We’re considering starting custom-key-only and adding routed billing (dedicated key + reconciliation) in a later phase, so we’d like to understand the plan prerequisites for each path. Thanks!

Happy to continue here or by email if that works better. Thanks!

Request to become an Inference Provider: Umans AI (GLM 5.2 + Kimi K2.7-Code)

Hi HF team 👋

We'd like to list Umans AI as an Inference Provider. We serve open-weight models on our own infrastructure, starting with:

• zai-org/GLM-5.2
• moonshot/kimi-k2.7-code (Kimi K2.7-Code)

On our side:

• HF org umans-ai is on the Team plan.
• We expose an OpenAI-compatible API (/v1/chat/completions, and /v1/models already serving pricing + context_length), which is the surface HF integrates with, plus an Anthropic Messages-compatible API (/v1/messages).
• We've implemented the billing callback (it maps each request id to its cost in nano-USD) and emit a unique Inference-Id response header, per the provider docs.
• Ready to open the huggingface.js provider PR and register model mappings as soon as our org is enabled server-side.

Could you point us to the next step to enable the org for the Model Mapping API, and the right contact for the commercial/payout setup?

Public API base: https://api.code.umans.ai
Contact: contact@umans.ai

Thanks!