Welcome Fireworks.ai on the Hub 🎆

Following our recent announcement on Inference Providers on the Hub, we're thrilled to share that Fireworks.ai is now a supported Inference Provider on HF Hub!

Fireworks.ai delivers blazing-fast serverless inference directly on model pages, as well as throughout the whole HF ecosystem of libraries and tools, making it easier than ever to run inference on your favorite models.

Among others, starting now, you can run serverless inference to the following models via Fireworks.ai:

• deepseek-ai/DeepSeek-R1
• deepseek-ai/DeepSeek-V3
• mistralai/Mistral-Small-24B-Instruct-2501
• Qwen/Qwen2.5-Coder-32B-Instruct
• meta-llama/Llama-3.2-90B-Vision-Instruct

and many more, you can find the full list here.

Light up your projects with Fireworks.ai today!

How it works

In the website UI

Search for all models supported by Fireworks on HF here.

From the client SDKs

from Python, using huggingface_hub

The following example shows how to use DeepSeek-R1 using Fireworks.ai as your inference provider. You can use a Hugging Face token for automatic routing through Hugging Face, or your own Fireworks.ai API key if you have one.

Install huggingface_hub from source:

pip install git+https://github.com/huggingface/huggingface_hub

Use the huggingface_hub python library to call Fireworks.ai endpoints by defining the provider parameter.

from huggingface_hub import InferenceClient

client = InferenceClient(
    provider="fireworks-ai",
    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)

from JS using @huggingface/inference

import { HfInference } from "@huggingface/inference";

const client = new HfInference("xxxxxxxxxxxxxxxxxxxxxxxx");

const chatCompletion = await client.chatCompletion({
    model: "deepseek-ai/DeepSeek-R1",
    messages: [
        {
            role: "user",
            content: "How to make extremely spicy Mayonnaise?"
        }
    ],
    provider: "fireworks-ai",
    max_tokens: 500
});

console.log(chatCompletion.choices[0].message);

From HTTP calls

Here's how you can call Llama-3.3-70B-Instruct using Fireworks.ai as the inference provider via cURL.

curl 'https://router.huggingface.co/fireworks-ai/v1/chat/completions' \
-H 'Authorization: Bearer xxxxxxxxxxxxxxxxxxxxxxxx' \
-H 'Content-Type: application/json' \
--data '{
    "model": "accounts/fireworks/models/llama-v3p3-70b-instruct",
    "messages": [
        {
            "role": "user",
            "content": "What is the meaning of life if you were a dog?"
        }
    ],
    "max_tokens": 500,
    "stream": false
}'

Billing

For direct requests, i.e. when you use a Fireworks key, you are billed directly on your Fireworks account.

For routed requests, i.e. when you authenticate via the hub, you'll only pay the standard Fireworks API rates. There's no additional markup from us, we just pass through the provider costs directly. (In the future, we may establish revenue-sharing agreements with our provider partners.)

Important Note ‼️ PRO users get $2 worth of Inference credits every month. You can use them across providers. 🔥

Subscribe to the Hugging Face PRO plan to get access to Inference credits, ZeroGPU, Spaces Dev Mode, 20x higher limits, and more.