OpenAI-compatible API.
Same SDK. Lower cost.
Use the same SDK and endpoints. Access multiple SOTA OSS models with predictable pricing.
Everything you need to ship
OpenAI-compatible API with predictable pricing and multi-model access.
Drop-in compatible
Full OpenAI API compatibility. Change your base URL and your existing integrations keep working.
Lower cost per token
Pay per token with clear per-model pricing, or use a plan for higher throughput.
Multi-model access
Open-weight and proprietary models — including vision and reasoning — available through one endpoint.
API key management
Create named keys per project. Revoke individually without disrupting other integrations.
Usage dashboard
Monitor token usage by model, track credits, and manage your plan from one place.
Low latency
Optimized inference pipelines with low time-to-first-token.
Switch in one line
If you've used the OpenAI SDK, you already know how to use CrofAI.
from openai import OpenAI client = OpenAI( base_url="https://crof.ai/v1", api_key="your-api-key", ) response = client.chat.completions.create( model="glm-5", messages=[{"role": "user", "content": "Hello!"}], ) print(response.choices[0].message.content)
Simple, predictable pricing
Start with per-token billing at no monthly cost. Upgrade to a plan when you need higher throughput.
Ready to build?
Switch by changing one line. Drop-in compatible with your existing OpenAI integrations.
Get Started