flo2

Flo2 is a developer-first LLM gateway, router and proxy . It connects your app to any AI model — OpenAI, Anthropic, Groq, Cerebras, DeepInfra and more — through a single API key, then automatically picks the cheapest and fastest model for every call. You bring your own provider keys. flo2 routes between them with zero token markup — you pay providers directly, no resale, no hidden fees. Key features: — Smart routing: pin a default model, restrict to a few, or open to your full key collection — Fallback chains: if the primary model errors, flo2 slides to the next fallback automatically — AI racing: fire multiple models in parallel, the fastest response wins — A/B testing: shadow new models against your live setup; a judge model scores which performs better — Full cost transparency: every call logs tokens, throughput and computed cost per provider Free during Beta.