Fireworks AI runs an inference platform for open generative AI models. Developers can call popular open language models through an API that is designed as a drop-in alternative to closed-model interfaces. The company also offers training and fine-tuning of models, and dedicated deployments with reserved capacity. The platform is delivered as a cloud service with serverless, on-demand and reserved tiers.
Fireworks AI, Inc. is a US company, and its privacy policy states that servers for the service are located in the United States. Support and engineering teams may work from the US. For transfers from the EEA and the UK, the policy refers to standard contractual clauses and to an appointed EU privacy representative. The page does not mention an EU hosting region or publish a data processing agreement. Because many of the models offered are openly licensed, organisations can run comparable models on their own infrastructure, but the platform itself is a US-hosted service.