Fireworks AI

Qwen team keeps shipping: Qwen3 Coder 480B is live on @FireworksAI_HQ - on par with Sonnet 4 for coding! 🤯 app.fireworks.ai/playground?mod… Quick impressions: • very strong agentic coding performance: SWEBench, Aider-Polyglot and other benchmarks are at the level of Claude Sonnet 4!…

We’re excited to share that our CEO, @lqiao, will be delivering a keynote at RAISE Summit 2025, taking place at the iconic Louvre in Paris. She’ll be speaking about Open Models and their critical role in shaping the next generation of AI agent development. If you’re attending…

During @FireworksAI_HQ Dev Day, we hosted a fireside chat between Malte Ubl, CTO of @vercel , and @lqiao, our CEO and co-founder, to dive deep into the future of developer workflows, GenAI infrastructure, and how AI is reshaping the software stack. A few key takeaways from the…

Three weeks ago, we started building an AI game engine. But some models kept making things look... sloppy. So we turned finding the best one into a game. In three weeks, that game grew to 35K+ users across 135 countries. Introducing @designarena_ai, the fastest-growing…

New Qwen3-235B-A22B-2507 is here, taking the spot as the SOTA non-reasoning open weights LLM from Kimi K2, just 9 days later and while being 4x smaller! 🤯 It’s amazing to have multiple open models at the level of closed providers, bridging the prior gaps for agentic use and…

Scaling agentic AI is tough — most teams struggle. @SentientAGI did it in 3 month with Fireworks. Sentient Chat launched with: - 15 agents - real-time search - multi-model support - 1.8M waitlist signups in 24h - 5M+ queries in days Read about the partnership here →…

I built an interactive visualization tool to understand how MuonClip helped with @Kimi_Moonshot 's K2 training! Try it yourself: …-app-644257448872.us-central1.run.app (1/11)

24 hours of non-stop stress-testing, Kimi-K2-Instruct cleared every benchmark in our playbook. 🚀 Now it’s LIVE on the Fireworks Serverless API It is the first open-weights SOTA 🔧 agentic tool-caller, holding its own on SWE Bench, Tau2 & AceBench. Same prod weights, zero infra…

If you’re building production-grade AI systems, function calling is fundamental. At Fireworks AI, we’ve been working with teams pushing LLMs from text generation toward agentic behavior- systems that plan, act, and reason with structured tools. Function calling is at the heart…

How do you assign rewards in tasks where “correctness” is subjective- like style, coherence, or creativity? Traditional reward functions struggle in open-ended domains. But in our latest blog, we show how using a stronger LLM as a judge offers a powerful alternative. With the…

We’re excited to co-host an evening at the AWS GenAI Loft in San Francisco, bringing together builders and experts to talk about scaling infrastructure for real-world GenAI applications. Our Field CTO, Shaunak Godbole, will be sharing how Fireworks AI is enabling low-latency,…

Fireworks is heading to AWS Summit NYC on July 16! After the event, join us and @superannotate for an exclusive post-summit happy hour. It’s your chance to connect with GenAI builders, innovators, and decision-makers. What to expect: ➡️ How to close the quality gap between…

MiniMax-M1-80k is now available on Fireworks! Why it matters: ➡️ Extreme long-context reasoning – Natively supports 1 million tokens (8× more than DeepSeek R1). Ideal for use cases like multi-file code refactoring, legal document analysis, and multi-turn agent workflows. ➡️…

🚀 MCP Support with Fireworks Responses API If you’ve tried wiring an LLM-powered agent into real systems, you already know the drill: - Endless boilerplate for calling tools and parsing tool-returns - Error-handling and retries that sprawl across your codebase - Latency…

At @FireworksAI_HQ Dev Day, we hosted a fireside chat with Sarah Sachs, Head of AI Engineering at @NotionHQ, and our co-founder Benny Yufei Chen, to explore how Notion’s AI journey evolved from personal productivity to enterprise-grade intelligence. Key takeaways from the…

In this fireside chat from @FireworksAI_HQ Dev Day, Adarsh Hiremath, CTO of @mercor_ai, joined our co-founder Benny Chen to unpack how AI models are transforming the way companies identify, evaluate, and hire global talent. Some key takeaways from the conversation: → Model…

In this fireside chat, Tony Wu, VP of AI at @perplexity_ai, joined Lin Qiao, our CEO of @FireworksAI_HQ, to explore what it means to run world-class research and infrastructure in tandem. A few key themes from the discussion: → Perplexity’s research team is hands-on with model…