Fireworks AI is a company that provides generative AI solutions to bridge the gap between prototype and production for businesses.

Fireworks AI

Fireworks.ai offers generative AI platform as a service. We optimize for rapid product iteration building on top of gen AI as well as minimizing cost to serve. 

https://fireworks.ai/careers

Cursor has trained a specialized model on the "fast apply" task, which involves planning and applying code changes. 

The fast-apply model surpasses the performance of GPT-4 and GPT-4o, achieving speeds of ~1000 tokens/s (approximately 3500 char/s) on a 70b model.

Read more about how Cursor built Fast Apply using the Fireworks AI Speculative Decoding API 👇 

https://lnkd.in/e9cPqFEx

Keeping up with our same-day ritual for hot new models at Wordware (YC S24) we release Gemma 2 9B from Google. 

Thanks to Fireworks AI for helping to make this happen, accelerated by my nudging at the AI Engineer World Fair. 

#llm #ai #google #gemma2

I'm excited to announce FireAttention v2: a breakthrough for long context inference. 

At Fireworks AI, we challenge ourselves to deliver high impact innovation  -- can we develop a dramatically more efficient serving for LLMs with RAG in a long context window with negligible quality tradeoff? 

We built our next generation runtime, FireAttention v2. When compared with vLLM, the Fireworks stack now serves Qwen2 and Mixtral (32k context window)  up to ~12x faster and ~8x higher throughput. 

For technical details, check our blogpost:

Join us in SoCal at the Multimodal AI hackathon! Use Fireworks to analyze and generate images and text to build the future of entertainment!

Excited to announce that our SDXL, Playground v2.5 and Segmind SSD image generation models are faster and less expensive!  

💨  0.8 seconds to generate a 30 step-image in our tests. Independently benchmarked to be the fastest! 
 💰 Prices reduced ~35% from $0.006 to $0.0039 for a 30-step image

AI:FIREWORKS

Fireworks AI