Question 1

How does Mimu pricing work?

Accepted Answer

Mimu offers simple monthly plans — Basic, Premium, and Max — plus pay-as-you-go billing for teams who prefer per-token pricing. Because every request is routed to the most cost-effective available provider, most workloads run up to 60% cheaper than going direct. No minimums on pay-as-you-go, and volume pricing is available for teams processing 10M+ tokens per month. Pay with Stripe globally or Razorpay in India.

Question 2

How quickly can I migrate from OpenAI?

Accepted Answer

Migration takes less than 5 minutes. Our API is OpenAI-compatible - just change your base URL and API key. We provide migration scripts for Python, Node.js, and popular frameworks. Your existing prompts and parameters work without modification.

Question 3

Which models and providers do you support?

Accepted Answer

Access 70+ models including GPT-5, Claude Opus 4.7 / Sonnet / Haiku, Gemini 3, Grok, Llama, Mistral, and Command R+. We add new models within 24-48 hours of release. Custom model endpoints and fine-tuned models are also supported.

Question 4

How do fallbacks and retries work?

Accepted Answer

Configure primary and backup models with automatic failover on errors, rate limits, or timeouts. Set fallback chains (e.g., GPT-5 → Claude → Gemini) with configurable retry logic. Transparent to your application - same response format regardless of which model responds.