Lumi API
Home
⌕
PricingDocsConsole
WorkbenchLLM BenchSign inSign up free
Home/Blog

Blog

Field notes on integration, billing and running API keys in production — no filler, and no price figures that go stale.

Latest posts

  • One base_url, many models: migrating to OpenAI-compatible routing

    Only three things change when you move OpenAI SDK code to an aggregation gateway: base_url, api_key and model. The order, five traps, a safe rollout.

    Oct 27, 2026·4 min read
    • Integration
    • Migration
    • OpenAI compatibility
  • Pay-as-you-go, per image, per second: estimating what one call costs

    Text bills per token, images per image, video per second. Formulas you can apply, why long threads get expensive, and four cost items people forget.

    Oct 20, 2026·3 min read
    • Billing
    • Cost
    • Estimation
  • Key management and rate limits: least privilege for a team

    One shared key cannot be attributed or revoked cleanly. Four least-privilege rules, the two rate-limit dimensions, and how to treat 429s as normal.

    Oct 13, 2026·3 min read
    • Security
    • Key management
    • Rate limits

Ready to make your first call?

Create a key, point your existing SDK at one base_url, and leave the rest of your code untouched.

Read the quickstart →See pricing
Lumi API

One endpoint to the world’s leading AI models. OpenAI-compatible and Anthropic-compatible — change one base_url.

Product
Pricing & model listConsolePlayground
Resources
QuickstartAPI referenceAgent integrationBillingBlog
Company
EnterprisePartner programContact supportTerms of servicePrivacy policy
© 2026 Lumi APIsupport@lumiapi.ai · Weekdays 9:30 – 21:00