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 !!

OpenRouter!

Hi everyone and first of all I really thank huggingface team for releasing this feature.
One another suggestion that I need is :
to give more granular parameters for the user to share the deployment configuration file (as an terraform hcl configuration) in order to define more optimum infrastructure for inference and let them define the inference model.

requesty.ai
groq

TypeError: InferenceClient.__init__() got an unexpected keyword argument 'provider'

Add RunPod as an inference provider!

Hyperbolic!

Runpod!

@llamameta please, make sure you use the latest version of huggingface_hub

pip install --upgrade huggingface_hub

Hello. Congratulations on great feature. How we can proceed on adding deepinfra.com to the providers list?

That's a great news, and if you provide more measurable details on what's free inference with a small quota is about for signed-in free users, you'll build more trust around the community, hence more PRO subscribers!

requesty.ai !

Runpod is excellent.
Let's add it to inference provider list

runpod!

Runpod would be very welcome!

Yes, Runpod.io would be great indeed

@julien-c

is there a way to add https://nineteen.ai/ as a provider ? it allows free access to top models like Deepseek R1 & the llama family
there is also : https://chutes.ai/ which can allow users to deploy any model on demand

Can we Add https://avian.io as a provider @julien-c

Currently have the fastest inference on Nvidia hardware, as well as highest throughput, allowing users to deploy any model on demand

The inference API is not working for HF Inference (∞ requests).

My Settings:

The Issue:

ai/ml api - dudes have a lot of models from HF!

try Runware.

Runware plz – it's ~5x cheaper than other providers and still one of the fastest

Try Scaleway.

Runware is best kept secret in the industry

I can't use DeepSeek R1 with hf inference npm , got this error :

ChatButton.tsx:120 Error: Error: Model deepseek-ai/DeepSeek-R1 is not supported for task conversational and provider together
    at mapModel (@huggingface_inference.js?v=2d418915:235:11)
    at makeRequestOptions (@huggingface_inference.js?v=2d418915:158:13)
    at streamingRequest (@huggingface_inference.js?v=2d418915:434:31)
    at streamingRequest.next (<anonymous>)
    at chatCompletionStream (@huggingface_inference.js?v=2d418915:933:10)
    at chatCompletionStream.next (<anonymous>)
    at handleSendMessage (ChatButton.tsx:94:21)

from code :

    const client = new HfInference(hf_token);
for await (const chunk of client.chatCompletionStream({
                model: "deepseek-ai/DeepSeek-R1",
                provider: "together",
                messages: [{ role: "user", content: input }],
                temperature: 0.5,
                stream: true,
            })) {
...

Runware would be great. I can run all of my models with them.

runware currently costs the same as just the electric power to run a 4090 in germany xD

Add Novita AI pls!!

I prefer Novita AI. Deepseek-r1 is currently more stable on Novita than on the official platform.

Novita AI!!

fireworks, they are super-fast

segmind.com is the best, cheap and fastest

Would love to see beam.cloud!

STOP LYING!!! THE FCKIN LAW SAYS THAT I'VE 1000 REQUESTS PER DAY AS A FREE USER, AND 20'000 PER DAY AS A PAID ONE... I'D JUST BECAME A PRO MEMBER, I'D PAY 9 FCKIN DOLLARS AS A POOR GUY, AND AFTER 400 REQUESTS OUT OF TWENTY THOUSAND(!!!), USING ONLY HF INT, THE SYSTEM LIES THAT I'D EXCEEDED MY MONTHLY QUOTA!!! (LITTLE MATH: THE MONTHLY QUOTA IS 600 THOUSAND REQUESTS(!!!), I'D BARELY USED THE HALF OF THE FREE MONTHLY QUOTA THIS MONTH (10'000)...) I'D SEARCHED GOOGLE, ASKED GPT, IT SAYS NOTHING(!!!) ABOUT QUOTA CHANGE FOR HF API... IT'S MY LIVING(!!!) THAT I GENERATE PICS BY SD 3.5, POST IT ON DEVIANTART, AND THEN I SELL IT AS NFTS TO BUYERS... YOU'RE RISKING MY LIVING WITH YOUR SILLY LITTLE GAME, HUGGINGFACE!!! I LOVE THIS PLATFORM, IT HAS THE MOST LIBERAL OF ALL, AND FOR MONTHS, I'D NO PROBLEMS... REALLY, WHY YOU WANT A HF-GATE?!?! CAN I TRUST YOU??? OR MONEY IS NOW YOUR GOD TOO, LIKE ALL THOSE MONEY HUNGRY SH*T SITES??? PLEASE, END THIS NIGHTMARE NOW!!!

Are the use of the new Inference Providers through the HF proxy covered by the same privacy policy as Hugging Face directly? No logging beyond short technical reasons, and no use of input for training?

add https://nineteen.ai/ , it's free

Yesseerr, I'm happy that freedom is more important for you than money :)

Add RunPod

Not working with entreprise token in routed mode:

"You have exceeded your monthly included credits for Inference Endpoints. Subscribe to PRO to get 20x more monthly allowance."

Is there a page, where I can filter out models being supported by different providers? Currently one needs to go inside the model card, of every model to find out.

HuggingFace, please, provide as a documentation, that exactly tells us how much an Interference costs. (Example: Stable Diffusion HF Interference 0.001$/image etc.)

I have been using this new feature with different providers to run Flux and I really like it a lot! But today february 4 it changed from 20,000 daily inferences to $2 dollars montly (200 inferences). I know its fair, and I actually don't use 20,000 inferences, not even 20 daily. But the change from having 20,000 daily inferences to having 6 daily, is shocking, you could have make it gradually or not even giving 20,000 daily since the beggining. It's hard to asimilate that sudden change 😭

HuggingFace, please, provide as a documentation, that exactly tells us how much an Interference costs. (Example: Stable Diffusion HF Interference 0.001$/image etc.)

this is in progress, currently working on it

@Moibe do you use HF-inference, or external providers?

I have been using this new feature with different providers to run Flux and I really like it a lot! But today february 4 it changed from 20,000 daily inferences to $2 dollars montly (200 inferences). I know its fair, and I actually don't use 20,000 inferences, not even 20 daily. But the change from having 20,000 daily inferences to having 6 daily, is shocking, you could have make it gradually or not even giving 20,000 daily since the beggining. It's hard to asimilate that sudden change 😭

Right, it was a bit shocking. I would always like to use the HF-inference, (as my first choice). Initially I thought that those models that have HF-inference will be consuming from the existing quota (of 20,000). Only the 3rd party ones we have leverage of $2 for pro members. I agree, that 20,000 is bit too much, but the change is rather drastic.

let me run the numbers, but for HF-inference the change should not be that drastic (most HF-inference requests are priced very very cheaply especially of course CPU based models)

The 20,000 daily figure was a bit unrealistic given it was "best-effort" meaning the rate of failures was quite high. Note that from now on, we exclude failing requests from those counts (we didn't up till now)

Any inference provider you love, and that you'd like to be able to access directly from the Hub?

I tested it and I love it, super easy!

The first question from my company was:
With inference providers can we setup something like private endpoints?

@Moibe do you use HF-inference, or external providers?

@julien-c yes I was using the hf-inference and usage was deducted from the 20,000. And about other providers, it allowed me to use them totally free (without substracting from the 20,000) and with a limited quota, meaning that I was able to make some usage of fal-ai until it messages that quota was excedded, then I used together, sambanova, etc. It was great beause it allowed me to test all provider even without having an account. I understand the change, and I think prices are still reasonable. I was just a shocking change, but I'm fine. I even got an account on fal.ai after the free test and I'm satisfied with the service amd the ease of use all from hf interface.

Let's add nineteen.ai PLEAAASE

With inference providers can we setup something like private endpoints?

@levalencia i don't think so but have you looked into Inference Endpoints? (it's dedicated instances)

THE HF INT API IS DENYING MY SERVICE BARKIN' "FAILED TO FETCH" (ILLEGAL EXCUSE!!!) (USED 0.20 OUT 2$ AS PRO)!!! SUPPORT, END THIS MADNESS NOW!!! TOMORROW, COURT AND PRESS, IF DON'T!!!

Subject: Add new provider to the Inference Endpoint.

I work at https://docs.predictionguard.com/home/getting-started/welcome and would like to explore if it is possible to add Prediction Guard to the list of inference providers on Hugging Face. Let me know how to go about this.

"ai/ml api - dudes have a lot of models from HF!"

We just launched our inference product here at io.net called IO Intelligence! Would love to have this plugged into HF as a new inference provider. Currently free to use until 500k token/day limit but we'll have the lowest price per token for the models we host once we roll out full pricing.

Feel free to try it out here: https://ai.io.net/ai/models

I'm really confused on this, it seems like things got a lot worse? Am I understanding that correctly? I'm not really sure how much an individual call would cost or how it can be calculated for a particular endpoint. Maybe I'm just confused but it seems like this announcement is essentially just saying pro subs had a pretty major feature removed. I've never really even used the inference features, but this reads like they are essentially gone now, right?

How can an Enterprise pay for Inference Credits ?
I'm a member of an Org with Enterprise Hub subscription, and I set up a payment method in the billing section of my org account. It seems that an access token can't be created by an Org, so I created an access token using my account, and it does show up in the access tokens of my Org. However, when using Inference providers, it seems to be only consuming my Inference Credits, and not trying to use the payment method of the Org. Is there a way to use Inference providers will billing the Org ? Thanks.

The Python docs for Inference Providers on hub seems to have a small error. I was going to propose a PR, but I couldn't find the source - is this available somewhere?

Steps to reproduce:

1. Go to a model page eg. https://huggingface.co/bigcode/starcoder2-15b
2. Click Deploy
3. Click Inference Providers
4. Click Python (deep link: https://huggingface.co/bigcode/starcoder2-15b?inference_provider=hf-inference&inference_api=true&language=python)

You'll now see

from huggingface_hub import InferenceClient

client = InferenceClient(
    provider="hf-inference",
    api_key="hf_xxxxxxxxxxxxxxxxxxxxxxxx"
)

result = client.text_generation(
    model="bigcode/starcoder2-15b",
    inputs="Can you please let us know more details about your ",
    provider="hf-inference",
)

print(result)

The problem is that the provider parameter should not be passed in via client.text_generation. It's correctly passed above when the client is instantiated via InferenceClient, but should be removed from client.text_generation.

@julien-c really you should include https://runware.ai as an inference provider.

It is by far the fastest and the cheapest provider with the most comprehensive API for making visual content: just have a look at this awesome documentation: https://runware.ai/docs/en/image-inference/api-reference

A FLUX DEV image is generated in 1.9 seconds with a price of 0.0038 (yes, 3 zeros, almost 10x cheaper than anybody else) , there are 300k+ models on the platform

and they are backed by A16Z, Lakestar and Midjourney

RunPod! The most flexible option out there.

Can you add DeepInfra?

Is there a way to programmatically list the providers and the models available from each of them?

When I try, the following, it fails (I have upgraded to latest huggingface_hub version)

client = InferenceClient(provider="nebius",)
client.list_deployed_models()

Thanks

let me run the numbers, but for HF-inference the change should not be that drastic (most HF-inference requests are priced very very cheaply especially of course CPU based models)

Really? 9$ for 200 messages in chat is a very cheap?

Hello @julien-c ,
I kindly request an increase in the quota limit for API Inference endpoints. As a student with a deep passion for exploring technology through these models, Hugging Face has always been an invaluable resource for me. Unfortunately, the current limits are proving to be a significant barrier, even after investing in the Pro subscription, as they restrict my ability to explore and develop projects in a cost-effective manner.
I believe many students and enthusiasts who love the Hugging Face platform would greatly benefit from a more generous quota. Your consideration in making the platform more accessible would be greatly appreciated.
Thank you for your time and support!

It is great to have the inference providers that allowing people to use the model easily. And I know other model providers also want to cooperate with Hugging Face to create easy, stable, and affordable services to the community. How can they add their api to the Hugging Face? Is there any instruction to do it? Thank you.

This new system is really bad for the users (us/me). Instead of 200.000 api requests like before I am not even getting 5.000 api requests. Already hitting max usage. If you are planning to give us far less credit then we are actually paying for the membership, I would consider not taking the membership and just pay for the spent credits. It is way cheaper. Now I feel like I am paying over 4 times more then the provided services are worth.

Add Runway please

@OFT so we can take a look can you share (here or in DM) which model(s) you're mostly hitting?

And i'm assuming it's on provider=hf-inference?

@OFT so we can take a look can you share (here or in DM) which model(s) you're mostly hitting?

And i'm assuming it's on provider=hf-inference?

@julien-c Thank you for having a look into this.

I am mostly using models that are:

• IMG 2 TXT
• TXT 2 IMG (Flux, SD 3.5, own models, ....)
• TXT 2 TXT

I don't know where I can verify provider=hf-inference. Could you give me a CURL example with this option? (not phyton please)
However I do have the following billing information:
Mar 1 - Mar 31 | HF Inference API (4,585) | $2.02
Feb 1 - Feb 28 | HF Inference API (17,156) | $1.72
It is unclear what happend between the two months since I don't think I changed something?

Extra: After noticing the problem, I also jumped to "https://huggingface.co/settings/inference-providers" and disabled everything except the HF Interference API.

Please add kluster.ai!!

@julien-c please give us prices per request for each model using hf-inference. The question is because the prices for different models vary widely, and I want to know what models are expensive to manage my bills. At least detail view in billing page.

I had cancelled my OpenAI subscription in favor of the HF Pro subscription. Now I realize you took the inference away. I will be cancelling my Pro subscription because frankly I don't understand who asked for this. If I want to use other providers I will use other providers, why would I route it through HF? There is no value proposition whatsoever. Anyone doing production loads will get their own provider anyways.

I would have accepted a price increase or a decreased request limit in the Pro plan over this bad business move. I don't see a point in the Pro subscription anymore, why do you keep it?

@julien-c let's goo for nineteen.ai / chutes.ai 🔥🔥🔥

Many users are complaining because of prices, let's add these two and enable real free open source for everyone

can i add more money beside the 2$ limitation per month?
since the pay as you go dosent work yet
what are my options?

Hello everyone. I would like to know if there is any document that states how much a single image generation costs with the model "Stable Diffusion 3.5 Large", thanks in advance!
I am trying to estimate how much I would have to pay for N generations.

I would like to see Lambda as a provider available in the list!
https://lambdalabs.com/

OMG: "Model stabilityai/stable-diffusion-3.5-large-turbo does not exist"...

Please include kluster.ai as well! Thanks!

Important note about Pay-as-you-go (= PAYG) for Inference Providers: https://huggingface.co/posts/julien-c/158943939527784

For Inference Providers who have built support for our Billing API (currently: Fal, Novita, HF-Inference – with more coming soon), we've started enabling Pay as you go (=PAYG)

What this means is that you can use those Inference Providers beyond the free included credits, and they're charged to your HF account.

Would love to have kluster.ai added - thanks!

I've been using kluster.ai and it is great, I would definitely like to see it added, thank you!

Hey folks we were recommended to post that Prodia (https://docs.prodia.com/) should become an inference provider as well because of how strongly we rank on Artifical Analysis for image gen models

Since yesterday, the ZeroGPU version of SD3.5 Large Turbo freezes out with ERROR (All the time...) Please Support, fix this problem as soon as you can...

@gulyasdavid1999 this is unrelated to Inference Providers though

I get a token 403 unexpected error: 403 Forbidden: This authentication method does not have sufficient permissions to call Inference Providers on behalf of user XXX

Code to reproduce :

from huggingface_hub import InferenceClient
client = InferenceClient()
client.list_deployed_models("text-generation-inference")

question around enterprise accounts:

• each user gets a $2 quota?
• in order to apply the quota they have to use a key from their ent account and the X-HF-Bill-To header?
• how much does a call cost? I can't find any documentation...

Would be great to add Clarifai to the list. The platform is vendor-agnostic, supporting AWS, GCP, Vultr, and Oracle. We are planning to add a lot more providers soon.

Our company wants to provide some private models.
Is it possible in Model Mapping [ https://huggingface.co/docs/inference-providers/en/register-as-a-provider#3-model-mapping-api ] to have hfModel as a "stub" only and providerModel as a real model?

Just signed up with HF and had some questions for the general community to help us get started. We plan to use the Cerebras Inference Provider using direct calls rather than routing through HF itself.

With a Pro subscription, are there any limits to token usage or queuing constraints when using a custom API key and direct calls? The free tier on Cerebras did have such constraints.

Thanks in advance

Hey all, I'd like to make nCompass (https://docs.ncompass.tech/api-reference/quickstart) an inference provider on HF. We build GPU optimizations to be able to support an API without rate limits by maximizing GPU utilization. I would really appreciate it if someone could help us with the process of becoming an inference provider.

Hi I have a problem using Smoleagents HfApiModel Inference, I noticed that even though I belong to an enterprise organization, the inference API uses the credits of my free account

and not those of the organization

yet I read here (https://huggingface.co/docs/inference-providers/en/pricing)
that it should automatically use those of the organization.
I wrote even here (https://discuss.huggingface.co/t/hugging-face-payment-error-402-youve-exceeded-monthly-quota/144968/10?u=alexman83) and it seems there is a bug...
How we can solve it? For our company. is important use that service...

Thank you for you help!

With a Pro subscription, are there any limits to token usage or queuing constraints when using a custom API key and direct calls? The free tier on Cerebras did have such constraints.

@sh8459131 When using a custom key, requests are forwarded to Cerebras directly so their limits will apply

@alexman83 can you share some sample code you're using? We might need to update smolagents to expose the new bill_to parameter. cc @albertvillanova for viz

@julien-c of course!
Thanks!

from smolagents import CodeAgent
from extraction_smolagents.custom_tools import CSVRetrieverTool
from smolagents import HfApiModel

from huggingface_hub import login
login()

prompt_template = """
# Prompt per Analisi e Estrazione di Topic a partire da una richiesta dell'utente
Sei un esperto analista di contenuti televisivi. Devi estrarre una lista di topics a partire da una richiesta dell'utente nel seguente modo

## Passaggio 1: Analizza la richiesta dell'utente
- Comprendi dalla richiesta dell'utente, indicata dopo la parola 'richiesta', quali sono i topic di suo interesse

## Passaggio 2: Confronto tematiche estratte con quelle fornite
- Leggi il file csv 'topics_info.csv' contente come informazioni il nome del topic (colonna name), le parole rappresentative (colonna representation) e i documenti rilevanti (representative_docs)
- Confronta le parole rappresentative dei vari topic con le tematiche estratte al Passaggio 1 e memorizza solo le righe dei topic che soddisfano questo requisito
- Adesso analizza per i topic memorizzati al passo precedente i documenti rappresentativi e verifica quali siano simili ai temi estratti dalla richiesta dell'utente al Passaggio 1 e memorizzali

## Passaggio 3: Generazione dell'output
Genera un file json contenente:
- la lista dei topic estratti usando il valore della colonna name
- la motivazione per cui sono stati scelti

Organizza il file json come nel segunete esempio:

json
{
    "topics": [<topic_1>, <topic_2>, <topic_3>],
    "motivazione: <motivazione>
}
"""

retriever = CSVRetrieverTool()
llm_model = HfApiModel(model_id='Qwen/Qwen2.5-Coder-32B-Instruct')
agent = CodeAgent(
    tools=[retriever],
    model=llm_model,
    verbosity_level=2,
    additional_authorized_imports = ['pandas']
)

question = prompt_template + '\n' + "voglio tematiche musicali"
answer = agent.run(question)
print(f"Answer: {answer}")

This is the custom class for reading CSV

from smolagents import Tool
import pandas as pd

class CSVRetrieverTool(Tool):
    
    name = "csv_retriever"
    description = "Uses the provided path to access a csv file using pandas dataframe"
    inputs = {
        "path": {
            "type": "string",
            "description": "The path containing the filename of the csv to read",
        }
    }
    output_type = "string"

    def __init__(self, **kwargs):
        super().__init__(**kwargs)

    def forward(self, path) -> str:
        df = pd.read_csv(path)
        return df.to_string()
```

@alexman83 Merve (@merve ) opened https://github.com/huggingface/smolagents/pull/1260 which will expose the bill_to param in smolagents' InferenceClient 🔥

(you'll need to upgrade your smolagents version)

@julien-c We just started integrating us WaveSpeedAI http://wavespeed.ai as the inference provider. We provide blazing fast image and video generation services. Happy to see we are listed!

@julien-c why is wavespeed on the list? they serve like 2 models. runware is faster and better :)

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...

We have no rate limits at nCompass (https://app.compass.tech) with fast inference. Feel free to give it a spin if rate limits are something that's a bottleneck for you.

we want to be an inference provider and submit pr(https://github.com/huggingface/huggingface.js/pull/1424). Can anyone please help me see

I'm getting a little mad... The private servers aren't working since hours!!! The ZeroGPU spaces freeze out with "worker error" since WEEKS!!! And the dev team aren't urged to fix them... For least a week (during EASTER) the Stable Diffusion picture generators produced pics in APPLE II (80's) quality!!! At least they fix them, but we're losing our patience... OK, the (credit-based) interface version of SD3.5 is working, but that's a funny excuse to not fix the others... Actually, the admins are REQUIRED to fix these errors by law... Yes, you can say that WE may fix it, but many of us aren't senior programmers, and in fact, it's forbidden for average users to fix the API, as the problem is (almost) never local... So, I officially demand the developers to fix the HF API (affecting the spaces) by tomorrow morning (EU time)!!! I really don't want to file a case against you in court...

I am sad, too 🥹💔💔

Hi @julien-c , we at Dat1.co want to be an inference provider and have submitted a PR for the initial step. Could anyone help us move forward? https://github.com/huggingface/huggingface.js/pull/1460

Please add FPT AI Inference http://marketplace.fptcloud.com/
It's a newcomer but already has hundreds of users, thanks to its speed, stability, and competitive price.

black-forest-labs/FLUX.1-schnell is not properly working under HFInference provider, is this going to be permanent or there is an issue?

Actually it seems that HFInference is not working at all, do we need to use models now only via external providers????

The message for any HFInference is: "Our latest automated health check on this model for this provider did not complete successfully." Is this temporary or HFInference won't process certain models anymore, or maybe it is a bug?

Hi Team,

I want to register as a Inference Providers, Can you please suggest us the way forward process.

Thanks
Cyfuture

Dear Hugging Face Team,

Greetings from Cyfuture AI!

We are reaching out to explore a potential collaboration with Hugging Face. As a rapidly growing enterprise-grade AI solutions provider, Cyfuture AI offers robust and scalable inference capabilities powered by high-performance GPU infrastructure.
We would be keen to join Hugging Face as an official inference provider to support model deployment and inference workloads for your global community. We believe this integration would bring mutual value—enhancing access to affordable, high-speed inference while expanding our reach within the AI ecosystem.
Please let us know the next steps or any prerequisites required to move forward with this partnership.
Looking forward to your response.
Regards,

Cyfuture.ai
Email us at - info@cyfuture.ai

Hi Hugging Face Team,

We are from GmiCloud (https://inference-engine.gmicloud.ai). We want to be an inference provider on Hugging Face. At gmicloud, we focus on LLM inference optimizations. We started to follow the instruction at https://huggingface.co/docs/inference-providers/register-as-a-provider#register-the-provider. While it needs to reach out first. We’d greatly appreciate any guidance or support from the community on how to move forward with becoming an official inference provider on the platform.

Thanks in advance!

Gmi Cloud AI
Email us at - yijie.h@gmicloud.ai

Dear Hugging Face Team,

We're reaching out from Swarmind.ai, a high-performance AI infrastructure company, to express interest in becoming an official inference provider on your platform.

We offer scalable, GPU-powered inference optimized for production workloads, and believe this integration would benefit both communities.

Let us know the next steps to move forward.

Best,
Swarmind Team
ceo@swarmind.ai

Dear Hugging Face Team,
We're reaching out from metay.ai , a high performance AI infra company. We are super excited in becoming an official inference provider.
Pls let us know the next step to move forward.
Best,
Sam
my company email: samli@blockfish.io

CometAPI - Hugging Face Inference Provider Application

Dear Hugging Face Team,

We're reaching out from CometAPI to express our strong interest in becoming an official inference provider on the Hugging Face platform.

About CometAPI

CometAPI is a comprehensive AI platform that provides unified access to over 500 cutting-edge AI models through a single, powerful API. We serve as a bridge between developers and the latest AI capabilities, offering seamless integration with minimal complexity.

Key Highlights:

• 🚀 500+ AI Models: Extensive collection including GPT-5, Claude, GLM-4.5, Qwen3-Coder, Kimi K2, Grok 4, and many more
• 🎨 Multi-Modal Support: Text generation, image creation, video production, music composition, and audio processing
• ⚡ Unified API: Single endpoint for all AI services with OpenAI-compatible API format
• 🔄 Real-time Updates: Instant access to newly released AI models
• 💰 Competitive Pricing: Cost-effective solutions with flexible usage plans
• 🛡️ Enterprise Ready: Robust infrastructure with enterprise-grade security and reliability

Why Partner with CometAPI?

1. Comprehensive Model Coverage: We provide access to the industry's most extensive collection of AI models, including many that are not available through other providers.

2. OpenAI API Compatibility: Our API strictly follows OpenAI specifications for LLMs and VLMs, making integration seamless and straightforward.

3. High-Performance Infrastructure: Built on scalable, enterprise-grade infrastructure designed to handle production workloads with low latency.

4. Global Reach: Serving developers worldwide with reliable service and comprehensive documentation.

5. Active Community: Growing ecosystem with active Discord community and comprehensive developer support.

Technical Capabilities

• API Compatibility: Full OpenAI API compliance for chat completions and text generation
• Supported Tasks:
  • Conversational AI (text-generation, image-text-to-text)
  • Text generation
  • Multi-modal capabilities
• Authentication: Bearer token authentication system
• Billing Integration: Ready to implement request ID tracking and cost reporting API
• Rate Limiting: Robust rate limiting and quota management systems

Business Information

• Company: CometAPI
• Website: www.cometapi.com
• API Documentation: api.cometapi.com/doc
• Console: api.cometapi.com/console

Contact Information

Primary Contact: devpartner@cometapi.com

Additional Resources:

• Discord Community: Join our Discord
• Twitter: @cometapi2025
• GitHub: github.com/CometAPI
• Support Email: support@cometapi.com

Next Steps

We are fully committed to following the complete integration process outlined in your documentation:

1. ✅ Prerequisites: We have OpenAI-compatible API and understand the task requirements
2. 🔄 JS Client Integration: Ready to submit PR to huggingface.js repository
3. 🔄 Model Mapping API: Ready to register our model mappings once enabled
4. 🔄 Billing Implementation: Can implement the required billing endpoint API
5. 🔄 Python Client Integration: Ready to submit PR to huggingface_hub
6. 🔄 Documentation: Ready to create comprehensive provider documentation

Value Proposition for Hugging Face Users

By integrating CometAPI as an inference provider, Hugging Face users will gain:

• Access to 500+ AI models through a single provider
• Competitive pricing with transparent cost structure
• High reliability and low-latency inference
• Regular updates with the latest model releases
• Comprehensive multi-modal capabilities
• Enterprise-grade support and SLA commitments

We believe this partnership will bring significant value to the Hugging Face community by providing broader model access and enhanced AI capabilities.

We would greatly appreciate the opportunity to discuss this partnership further and are ready to begin the technical integration process immediately upon approval.

Thank you for considering our application. We look forward to your response and the possibility of collaborating with the Hugging Face team.

Best regards,

CometAPI Development Team
devpartner@cometapi.com
www.cometapi.com

Ready to revolutionize AI accessibility together? Let's make it happen! 🚀

Partnering on Inference – Qubrid AI (https://platform.qubrid.com)

Hi Hugging Face team and community, 👋

I’m with Qubrid AI, where we provide full GPU virtual machines (A100/H100/B200) along with developer-first tools for training, fine-tuning, RAG, and inference at scale.

We’ve seen strong adoption from developers who want dedicated GPUs with SSH/Jupyter access - no fractional sharing, plus no-code templates for faster model deployment. Many of our users are already running Hugging Face models on Qubrid for inference and fine-tuning.

We’d love to explore getting listed as an Inference Partner with Hugging Face, so that builders in your ecosystem can easily discover and run models on Qubrid’s GPU cloud.

What would be the best way to start that conversation? Is there a formal process for evaluation?

Looking forward to collaborating 🙌

Shubham
Team Qubrid

Dear Hugging Face community,

I'm reaching out from Runware. We provide a high-quality media through a fast, affordable API. We are excited in becoming an official inference provider on Hugging Face Hub.

We are already on the team plan. But still unable to post model mapping as we keep on getting Cannot POST /api/partners/runware/models error.

Kindly let us know the next step to move forward.

Best,
Sirshendu
email: sirshendu.ganguly@runware.ai

Hey there all,

We're Bytez, and we're the largest inference provider on the internet! (We offer inference for 170k+ models.)

We stepped through the integration guide and upgraded to a team account, but are encountering the same issue @sirshendu1997 is encountering.

We also reached out to enterprise-customer@huggingface.co but have not received word back after a few days, so we're trying this avenue.

Please let us know what we need to do to get our integration working. We'd love to be a provider for HF!

@julien-c

PublicAI and Scaleway have been added! 🥳

PublicAI and Scaleway have been added! 🥳

We'd also love to get listed, any help will be appreciated. (https://platform.qubrid.com/)

Hi,

I'm Charles from Boost Run. We own our compute and operate our data centers, running large clusters equipped with Blackwell, Hopper, and other model GPUs.

Back in 2022-24, I was one of the early engineers on Amazon Bedrock. At the time, the push for dedicated GenAI inference providers was still fairly new. I had a front-row seat to what worked, and what became fragile as things scaled. Our goal is to build a lean and efficient machine, processing the most important workloads in a secure and highly available manner.

I'll be going through the onboarding docs over the coming weeks and am excited to get more familiar with everything. Hopefully this can be the beginning of a strong partnership.

PS. AI is a better writer than myself 😀

Also feel free to reach out to me cr@boostrun.com or can invite to our slack org

About us

Hi Hugging Face team, 👋

I’m with GPT Porto (https://www.gptproto.com/), an AI API platform focused on providing safe, stable, fast, and affordable inference for developers and enterprises. With just one API key, our users can access most mainstream models, including Hugging Face models, while enjoying reliable infrastructure and cost efficiency.

We’ve seen strong adoption from teams who value predictable performance, security, and competitive pricing for both experimentation and production workloads. Many of them already integrate Hugging Face models through GPT Proto to streamline deployment and reduce costs.

We’d love to explore becoming an official Inference Provider on Hugging Face, so that more builders in your ecosystem can benefit from a secure, high-performance, and budget-friendly option for model inference.
Looking forward to collaborating!

Contact us: gptproto@venuslondontechnology.co.uk

Best regards,
Team GPT Proto

@julien-c

Hi Julien would I be able to reach out to you or someone from HF about the API Key management and Billing?

https://huggingface.co/docs/inference-providers/en/register-as-a-provider#4-billing

Thanks,
Boost Run

Prev: https://huggingface.co/spaces/huggingface/HuggingDiscussions/discussions/49#68d564aa7a7e392a927c11f6

Would be great to have Simplismart to the list!

Hi!

Would be great if Snowcell could be added to the list. We build complete inferencing solutions from the ground up.

I couldn't find a specific contact point to reach about this, but for any questions we are available at contact@snowcell.io.

Best Regards

Hello,

At FAIM, we are building an inference platform for time-series foundation models: https://faim.it.com/.
All models we currently support are available on Hugging Face.

I would like to clarify whether it’s possible for us to become an inference provider on Hugging Face for time-series models.

Thank you and best regards,
Andrei
andrei.chernov@faim.it.com

Hello Gatewayz is ready for integation. Please email me at vaughn@gatewayz.ai

Hi Hugging Face team!

We're gcube (https://gcube.ai), a GPU sharing platform from South Korea. We make AI inference super affordable by connecting idle GPUs from cloud providers and even PC cafes across Korea - basically turning unused computing power into a distributed GPU network.

Our customers are seeing 55-70% cost savings, and we work with major Korean cloud partners like Naver Cloud, NHN Cloud, and KT Cloud.

We'd love to become an official Inference Provider on Hugging Face. Would really appreciate any guidance on the next steps!

Our HF org: https://huggingface.co/gcube-ai (Team plan subscribed)

Thanks!

Best,
Koo
Data Alliance (gcube)
jykoo@data-alliance.com

Hello Hugging Face team 👋

We’re from Simplismart.ai, a Series A startup backed by Accel, building a modular MLOps platform focused on high-performance inference. We’re currently exploring the process of listing our inference APIs as a provider on Hugging Face.

We’ve gone through the inference provider documentation and are preparing for the next steps, but before raising a PR, we’d appreciate some clarity around the billing flow, specifically:

Questions:

1. What is the expected delay between a successful inference request and Hugging Face calling the billing endpoint?
2. If we’re unable to return cost details within one minute when Hugging Face hits the billing endpoint, does Hugging Face retry the request? If so, what’s the retry behavior?

We want to ensure our implementation aligns closely with Hugging Face’s billing expectations, so any guidance on the above would be very helpful.

Thanks in advance for the support! 🤗

-- Pratik Parmar
Developer Advocate @ Simplismart.ai
pratik@simplismart.tech

Hello Hugging Face team 👋

We are from AlphaNeural AI, an AI infrastructure and inference platform built around one core idea:
Developers should be able to deploy their models and make money from them easily.

AlphaNeural allows model creators to deploy their models on our infrastructure, set licensing or access rules, and monetize usage directly. These deployed models can then be exposed on Hugging Face, while AlphaNeural handles the inference layer with a strong focus on affordability, scalability, performance, and reliability.

We provide inference for LLMs, text, image, video, audio, and multimodal models, optimized for high throughput and low latency, powered by GPU infrastructure that is up to 70% cheaper than traditional cloud providers. This makes it possible for developers to price their models competitively while still earning meaningful revenue.

Our platform is designed to be:
• Developer friendly with simple deployment and API access
• Cost efficient for both creators and users
• High performance and production ready
• Scalable from experimentation to enterprise workloads

AlphaNeural also bridges Web2 and Web3, enabling both traditional subscription based monetization and on chain licensing for models, opening new revenue paths for creators without forcing any Web3 complexity on end users.

We would love to become an official Inference Provider on Hugging Face, enabling model authors to deploy and monetize their models through AlphaNeural, while Hugging Face users access them with fast, affordable, and reliable inference. To maintain high quality, AlphaNeural applies strict technical and operational standards for all deployed models, ensuring strong performance, stability, and production readiness across the platform.

From a technical perspective, we have already completed the full API integration according to Hugging Face’s inference provider specifications, prepared the required interfaces, and have the PR ready for review. We are now simply waiting for approval to move forward.

Happy to share technical details and follow the onboarding process.

We have two PRs ready for review:

https://github.com/huggingface/huggingface.js/pull/1907
https://github.com/huggingface/huggingface_hub/pull/3678

Please reach out to:
contact@alphaneural.io
ash.eryani@alphaneural.io

Looking forward to collaborating 🤗

Team AlphaNeural AI

Hi HuggingFace Team,

We would like to register Mokzu as an inference provider.

Provider Details:

• Organization: Mokzu (https://huggingface.co/mokzu)
• Website: https://mokzu.com

We have submitted PRs to both huggingface_hub and huggingface.js repositories:

• Python PR: https://github.com/huggingface/huggingface_hub/pull/3686
• JavaScript PR: https://github.com/huggingface/huggingface.js/pull/1920

Could you please enable the Model Mapping API for our organization and provide guidance on registering our provider?

Thank you!

Hi HuggingFace team! 👋

We'd like to register Latitude.sh as an inference provider.

About Latitude:

• Recently acquired by Megaport (ASX: MP1), a global Network-as-a-Service leader
• Operating 10,000+ physical servers and 1,000+ GPUs globally
• Combined platform spanning 1,000+ data centers in 26 countries
• Tier-3 data centers with 99.99% SLA

Technical Implementation:

• OpenAI-compatible API at https://api.lsh.ai
• Billing endpoint implemented (POST /api/billing/costs returning nano-USD)
• Inference-Id header in all responses (regular + streaming)
• Full support: tool calling, structured output (JSON mode), vision/multimodal, streaming

Submitted PRs:

• JS Client: https://github.com/huggingface/huggingface.js/pull/1927
• Python Client: https://github.com/huggingface/huggingface_hub/pull/3715
• Documentation: https://github.com/huggingface/hub-docs/pull/2180

HuggingFace Organization: https://huggingface.co/latitude-sh

Logo Assets:

• Light: https://logos-swart.vercel.app/latitude-light.png
• Dark: https://logos-swart.vercel.app/latitude-dark.png

Models Ready for Mapping:
┌──────────────────────────────────────────┬─────────────────────────┬────────────────┐
│ HF Model │ Provider Model │ Task │
├──────────────────────────────────────────┼─────────────────────────┼────────────────┤
│ Qwen/Qwen2.5-7B-Instruct │ qwen-2.5-7b │ conversational │
├──────────────────────────────────────────┼─────────────────────────┼────────────────┤
│ meta-llama/Llama-3.1-8B-Instruct │ llama-3.1-8b │ conversational │
├──────────────────────────────────────────┼─────────────────────────┼────────────────┤
│ Qwen/Qwen2.5-VL-7B-Instruct │ qwen-2.5-vl-7b │ conversational │
├──────────────────────────────────────────┼─────────────────────────┼────────────────┤
│ google/gemma-2-27b-it │ gemma-2-27b │ conversational │
├──────────────────────────────────────────┼─────────────────────────┼────────────────┤
│ Qwen/Qwen3-32B │ qwen3-32b │ conversational │
├──────────────────────────────────────────┼─────────────────────────┼────────────────┤
│ Qwen/Qwen2.5-Coder-32B-Instruct │ qwen2.5-coder-32b │ conversational │
├──────────────────────────────────────────┼─────────────────────────┼────────────────┤
│ deepseek-ai/DeepSeek-R1-Distill-Qwen-14B │ deepseek-r1-distill-14b │ conversational │
└──────────────────────────────────────────┴─────────────────────────┴────────────────┘
Please let us know if you need any additional information or if there are next steps we should follow.

Thank you!

Hey, I've been using inference providers from the CLI and kept running into the discoverability problem others mention here. I ended up building a small Rust CLI/TUI for it https://github.com/jadnohra/hf-providers (brew from tap). I understand it's not a solution, but happy to have eyes on it in case anyone finds it useful.

Hello, HF team / HF community

We're The Compute Factory, a multicloud AI inference platform from France 🇫🇷. We'd like to register as an Inference Provider.

Looking at the current HF inference providers landscape, we see that text-to-image and especially text-to-video are significantly underserved; only a handful of providers cover these tasks today. This is a gap we specifically intend to fill. Our GPU (+other accelerators) infrastructure is well suited for the heavy compute these models demand, and we plan to aggressively expand HF hosted models catalog for image and video generation alongside LLMs.

Our plan is to start with openai/gpt-oss-20B for chat completion, and Lightricks/LTX-2-19B for text to video generation as our first integrations; this lets us establish the end-to-end HF provider pipeline (API, billing, model mapping, PRs). From there, we'll rapidly deploy dozens of additional models across chat completion, text-to-image, and text-to-video.

We're technically ready and can start implementing the HF Inference Provider specs immediately.

HF org: https://huggingface.co/thecomputefactory

PR: https://github.com/huggingface/huggingface.js/pull/1988

We would love any guidance on next steps. Happy to jump on a call.

Best,
The Compute Factory team
chouki@thecomputefactory.com

Hello,

We are a small datacenter company in Germany and have some questions to become a inference partner.

Our questions:
• Can we offer a ocr model as an inference partner?
• How are the scaling limits regulated or can they be adjusted? We cannot serve an infinite number of users.
• Is it acceptable to start with only one model as a partner initially?

Thanks in advanced and best regards

Hello HuggingFace Team,

We would like to add TextCLF as an inference provider.

Provider Details:

Organization: TextCLF(https://huggingface.co/textclf-ai)
Website: https://textclf.com
We have submitted PRs to both huggingface_hub and huggingface.js repositories:

Python PR: https://github.com/huggingface/huggingface_hub/pull/3895
JavaScript PR: https://github.com/huggingface/huggingface.js/pull/2022
Could you please enable the Model Mapping API for our organization and provide guidance on registering our provider?

Email: contact@textclf.com

Thanks,
TextCLF

Hi HuggingFace team,

We’re the team behind Phosor, a GPU inference platform specializing in AI video generation. We’d like to express our interest in becoming a HuggingFace Inference Provider.

About Phosor

Phosor provides high-performance inference for the Wan model family in production, with full support for both Wan 2.1 and Wan 2.2, covering:

• Text-to-Video (T2V) and Image-to-Video (I2V) at 14B scale
• LoRA support — users can upload and apply custom LoRA weights at inference time
• Multi-resolution output — 480p, 720p, and 1080p with configurable frame counts
• On-demand model loading — our service can download and run models on demand
  We run dedicated GPU clusters with Ray-based job scheduling, and our RESTful API (/api/v1/inference/submit, /api/v1/inference/status/{request_id}, /api/v1/inference/result/{request_id}) follows established async patterns. Our infrastructure is designed for high availability with automatic failover, ensuring stable and reliable service for production workloads.

Pricing

We use a pay-per-use model, priced per frame:

LoRA Training: $0.003 per training step (~$1.20 for a typical ~400-step training run)

Lowest price: $0.0002/frame (480p) — significantly lower than comparable providers in the market.

Content Policy

• NSFW content generation: No
• Celebrity / public figure generation: No

Why We Think This Is a Good Fit

The Wan model family has strong presence on HuggingFace Hub, but there is currently no dedicated video generation inference provider. We’ve already invested significant engineering effort into optimizing these models for production — including ComfyUI-based pipelines, LoRA hot-loading, and efficient queue management. Our pricing is among the lowest in the industry — starting from just $0.0002 per frame at 480p, making AI video generation accessible to a much wider audience. Additionally, our platform is built for production-grade stability, with dedicated GPU clusters, automatic failover, and consistent uptime, ensuring reliable inference at scale. Integrating as a provider would give HuggingFace users one-click access to affordable, stable video generation directly from Wan model pages.

Integration Timeline

We’re ready to start immediately — ideally within 1–2 weeks. We’d appreciate guidance on:

• Technical requirements for API adaptation (endpoint specs, auth flow, billing integration)
• Documentation or reference implementations for the provider onboarding process
• Timeline expectations for review and approval
  Happy to align our API with whatever specification HuggingFace requires.

Contact

• Website: https://phosor.ai
• Email: hello@aoraki-labs.io
  Best,
  Aoraki Labs Team

Hi team,

I'm Ben, one of the founding engineers at relaxAI. We're a sovereign AI inference provider hosted on Civo infrastructure, the UK's leading sovereign cloud platform. We'd like to express our interest in becoming a HuggingFace Inference Provider.

About relaxAI

relaxAI provides high-performance, OpenAI-compatible inference for leading open source LLMs. We're currently serving models that are popular across your existing providers, including:

• gpt-oss-120b
• Kimi-K2.5
• Llama 4 Maverick

Our infrastructure runs on Blackwell GPUs through our NVIDIA partnership, and our throughput, latency, and token pricing are very competitive.

What Sets Us Apart

What makes us different from your current provider lineup is that we're fully UK sovereign — 100% UK data residency, processing, and legal jurisdiction. Sovereign AI is a growing priority for UK and European enterprises, particularly in regulated industries, and as far as we can tell there isn't currently a UK-domiciled provider represented in Inference Providers. We think that's a meaningful gap we could help fill.

Integration Readiness

We already have a Hub organisation at huggingface.co/relaxai and we're ready to get started immediately — JS client PR, model mappings, billing endpoint, the lot. Our API is fully OpenAI-compatible, so we'd expect the integration to be straightforward. Happy to provide API access for testing whenever suits.

We'd appreciate guidance on:

• Timeline expectations for review and approval
• Any provider-specific requirements beyond the standard onboarding docs

Contact

• Website: https://relax.ai
• API Docs: https://relax.ai/docs
• Email: ben@relax.ai

Best,
Ben
Founding Engineer, relaxAI

Hi Hugging Face team,

I'd like to introduce StrikeEngine — a new LLM inference provider launching in May 2026 — and express
interest in joining the Inference Providers program. I've read through the register-as-a-provider guide
in the docs and I'm planning to start with the JS client PR per Step 2, since our API is OpenAI-compatible
from day 1.

What we do :
Host AWQ-int4 quantized open-weight LLMs — Qwen, DeepSeek, MiMo, GLM, Llama, Mistral families — on a
hybrid compute architecture (Modal serverless for launch latency + own spot-GPU fleet across AWS/Azure/GCP
for sustained margin). OpenAI-compatible /v1/chat/completions, multi-region failover, SLA monitoring.
the Inference-Id header for the billing endpoint.

Our differentiation
First-responder onboarding of new Chinese-lab open-weight releases — targeting 24-72h from model drop to a
listed endpoint. The current provider lineup has excellent latency specialists (Cerebras, Groq,
Fireworks) and price leaders (Novita), but no one is explicitly focused on the "first-to-list
new-releases" angle. That's the gap we fill. Target catalog: Qwen3.5 / Qwen-Coder-Next, DeepSeek V4+,
MiMo, GLM-4.7+, Kimi-K2+, MiniMax-M2+ — with an automated pipeline detecting releases and onboarding them
within hours.

Why HF Inference Providers
HF is where the open-weights community lives. Our model focus (quantized open-weights from Chinese labs)
aligns exactly with what your community searches for daily. We complement rather than replicate existing
providers by closing the "waiting-for-someone-to-list-the-new-release" gap your buyers currently
experience.

Technical posture toward the 9-step integration

• Step 1 (Task API): OpenAI-compatible — should skip most task-specific work
• Step 2 (JS client PR): happy to open this first as the canonical entry
• Step 3 (Model Mapping): will upgrade to Team Hub plan when we reach this step
• Step 4 (Billing): implementing Inference-Id + cost lookup endpoint natively
• Steps 5-9: will follow in order with PRs to huggingface_hub and hub-docs
• We meet the <5s time-to-first-token requirement by design (hybrid serverless + NVMe-cached spot)

Where we are today

• Landing: https://strikeengine.dev
• Partnerships contact: partnerships@strikeengine.dev
• General: hello@strikeengine.dev
• Launch target: mid-end of May 2026

What I'd like to request

• Whether opening the JS client PR is the right first concrete action, or if there's a preferred order
• Any prerequisites I might have missed

Thanks!
Hicham Abadou
Founder, StrikeEngine

Hi Hugging Face team,

We're Oxlo.ai - an AI inference platform backed by Cyborg Network, featured in STL Partners' "Top 50 Edge Companies 2026." We serve 40+ open-source models (Qwen 3 32B, Llama 3.3 70B, DeepSeek R1, Whisper, SDXL, etc.) via fully OpenAI SDK-compatible APIs.

We'd like to integrate as an Inference Provider on the Hub. Our API is already OpenAI-compatible, so the technical integration should be straightforward.

Why Oxlo.ai is different: We use request-based pricing (flat fee per API call, not per token), which makes us complementary to existing token-based providers on the Hub.

We have 700+ active users across 100+ countries and are ready to begin the technical integration (huggingface.js PR, Python SDK PR, model mappings).

Could you connect us with the right person on your partnerships or developer relations team?

Contact:

• Website: https://www.oxlo.ai/
• Docs: https://docs.oxlo.ai/docs/
• Email: hello@oxlo.ai

Best,
Shashank MS, Oxlo.ai Team

Hey @julien-c we'd like to register as an inference provider and do some co-marketing too; please let me know way forward. We have already done the technical readiness work on our end; www.univars.space targeting to be the first inference provider partner from Africa.