Flatkey raises $10M to sell developers one API key for 100+ AI models
Flatkey announced a $10 million Series A and 10,000 developers since its July launch, pitching one key, one balance, and one invoice for models and tools across providers.
The API-key sprawl problem has a new funded answer. On September 29, Flatkey announced a $10 million Series A round, alongside news that more than 10,000 developers have used the platform since its July 2026 launch.
One key, one balance, one invoice
Flatkey's pitch is consolidation. Instead of separate accounts, credits, API keys, rate limits, and invoices for every provider, developers get one key and one balance that covers more than 100 official AI models and more than 1,000 AI tools. The models span OpenAI, Anthropic, Google, DeepSeek, Kimi, and GLM, plus image and video models like Seedance. The tools cover search, browsers, data enrichment, media generation, and actions.
The company stresses that every model call goes through an official provider endpoint; it does not self-host modified or quantized versions and label them as the original. New models arrive through official channels as they release, so developers skip setting up a new provider account each time something new ships.
The pricing angle
Because Flatkey buys upstream capacity in volume, it prices many APIs at 60 to 90 percent of providers' official list prices, with most models around 80 percent and deeper discounts on some as promotions. Subscription plans start at $10 per month, and pay-per-use credits cover both models and tools. It is a drop-in replacement for any OpenAI-compatible client: change the base URL and existing code works.
"AI development is becoming less about choosing one model and more about combining models, data and tools across text, image, audio and video," said founder Hunter Guo in the announcement. "If developers can reach all of that through one key, the platform stops being a convenience layer and starts to look like infrastructure. That is the company we are building."
The funding will go toward adding more official models and tools and scaling the routing infrastructure. The bet is straightforward: as agents combine models from half a dozen providers in a single workflow, the company that owns the billing relationship owns a chokepoint. Whether developers trust a middleman with their inference spend at production scale is the open question.
