Frequently Asked Questions
FreshHow do I get higher rate limits?
Contact the support team with details about your use case, expected volume, and any latency or throughput requirements. The team will evaluate your needs and adjust limits.
How do I handle rate limits?
Implement exponential backoff in your client code. When you receive a 429 (rate limit) or 503 (server overloaded) error, wait before retrying. Double the wait time with each successive retry.
How are tokens counted?
Tokens are pieces of text that the models process. A token is roughly 4 characters for English text. Both input (prompt) and output (completion) tokens count toward your usage.
Is the API OpenAI-compatible?
Yes. The Inception API uses the same endpoint structure and request format as the OpenAI API. You can use existing OpenAI client libraries by changing the base_url to https://api.inceptionlabs.ai/v1.
What models are available?
- Mercury 2 (
mercury-2) — Chat completions, tool calling, structured outputs. 128K context. - Mercury Edit (
mercury-edit) — Code autocomplete, apply edits, next-edit predictions. 32K context.
How much does it cost?
- Input: $0.25 per 1M tokens
- Cached Input: $0.025 per 1M tokens
- Output: $0.75 per 1M tokens
- New accounts get 10 million free tokens
What IDE integrations are supported?
Cursor, Cline, Roo Code, Kilo Code, Zed, OpenCode, and OpenClaw. See the Integrations section for setup guides.
What is diffusion mode?
Mercury 2 is built on a diffusion architecture. Setting diffusing: true (with stream: true) lets you visualize how noisy outputs are iteratively refined into polished text.
What is instant mode?
Setting reasoning_effort: "instant" enables near-real-time responses from Mercury 2, ideal for voice assistants, chatbots, and low-latency applications.