Fireworks.ai added as Inference Provider on Hugging Face Hub

AI Impact Summary

Fireworks.ai is now a supported Inference Provider on Hugging Face Hub, enabling serverless inference directly on model pages and across HF libraries. The integration demonstrates usage via the InferenceClient, supporting both Fireworks API keys and Hub routing tokens, with concrete examples for deepseek-ai/DeepSeek-R1 and llama-v3p3-70b-instruct, including the router.huggingface.co endpoint for chat completions. This change enables consolidated inference traffic through Fireworks.ai, with billing and performance that depend on whether requests are issued directly to Fireworks or routed through the Hub; teams should review credential management, model compatibility, rate limits, and any provider-specific credits or pricing (e.g., PRO credits).