One API. Smart Routing. Simple Pricing.
Flint auto-decomposes every prompt, routes each sub-task to the optimal model, and returns a synthesized answer. Pay for tokens, not complexity.
$5 free credit on signup · No credit card required
Starter
For side projects and prototyping
- ✓ Flint Smart Routing
- ✓ 100K tokens/month included
- ✓ 30 req/min
- ✓ All supported models
- ✓ OpenAI-compatible API
- ✓ Email support
Pro
For professionals and production apps
- ✓ Everything in Starter
- ✓ 1M tokens/month included
- ✓ 300 req/min
- ✓ Priority model routing
- ✓ Streaming support
- ✓ Function calling
- ✓ Priority support
Enterprise
For teams and high-volume workloads
- ✓ Everything in Pro
- ✓ 10M tokens/month included
- ✓ 1000 req/min
- ✓ Custom SLAs
- ✓ Volume discounts
- ✓ Dedicated support
- ✓ 99.9% uptime SLA
Frequently Asked Questions
What is Flint?
Flint is a smart routing engine. Send one prompt — it decomposes, dispatches each sub-task to the best model, and returns a single synthesized answer. You never pick a model.
Which models do you support?
Flint routes to DeepSeek V4, Qwen3.7, Kimi K2, GLM-5, MiniMax M2, and 30+ other models. You get optimal model selection without managing multiple API keys.
How does billing work?
Monthly plans include a token allowance. Overage is billed per 1M tokens at your plan's rate. $5 free credit on signup — no card required.
Can I access specific models directly?
Yes. While Flint auto-routes by default, set model="deepseek-v4-pro" (or any supported model) to bypass routing. Same pricing, same endpoint.
Ready to stop picking models?
$5 free credit. No credit card required. Start in 60 seconds.