Fireworks AI on LinkedIn: Fireworks - Fastest Inference for Generative AI (original) (raw)
Fireworks AI’s Post
DeepSeek V3, a state-of-the-art open model, is now available on Fireworks Serverless and Enterprise! 🥇 SOTA open model for coding and reasoning 🥇 Best performing open model on Chatbot Arena and WebDev Arena 🧠671B MoE parameters, 37B activated parameters DeepSeek V3 is great to use on Fireworks: 📜 Long context enabled - 131K context size ⚡ Up to 30 tok/s speeds and we’ll be making it faster! 💰 Priced at $0.9/M tokens This is an early release, we are continuing to optimize the model! We expect high demand at launch and we will be working to make the API endpoint faster. We have run some evals to ensure good model quality after quantization, and we will be running more comprehensive evals in the coming days. Start using the model in our playground or via API, and contact us for enterprise deployments. Playground: https://lnkd.in/g2BAa2m9Get started with API: https://lnkd.in/gHHJkWsmContact us for enterprise deployments: https://lnkd.in/gnKdVucH
fireworks.ai See more comments
``