Releases

Fireworks launched general availability of its Training API for managed distributed training and reinforcement-learning rollouts.

Fireworks reported serving more than 40 trillion tokens per day, over 95% from customer-specialized models.

Fireworks disclosed that annualized revenue run rate had surpassed $1 billion.

Fireworks AI raised a $1.505 billion Series D led by Atreides Management, Index Ventures, and TCV.

Fireworks and Microsoft made select open models available through serverless endpoints in Microsoft Foundry.

Fireworks AI raised a $250 million Series C co-led by Lightspeed Venture Partners, Index Ventures, and Evantic.

Fireworks released a declarative SDK for managing models, deployments, fine-tuning jobs, datasets, and serving strategy.

Fireworks launched its integrated platform for rapid tuning, evaluation, deployment, and online experimentation.

Fireworks added support for NVIDIA NIM microservices and compatible models on its inference platform.

Fireworks introduced f1 and f1-mini, compound-AI reasoning systems that interleave multiple open models at inference time.

Fireworks AI raised a $52 million Series B led by Sequoia Capital, bringing total funding to $77 million.

Fireworks released an open-weight Llama 3-based function-calling model supporting multi-turn and parallel tool execution.

Fireworks launched integrated LoRA fine-tuning with rapid deployment into its production inference platform.

Fireworks AI raised a $25 million Series A led by Benchmark.