Meida Developer Documentation

High-performance, sub-50ms intelligent LLM router with zero-downtime failover and 100% OpenAI SDK drop-in compatibility.

How It Works

Every incoming turn passes through a 3-step pipeline — classify intent, select the cheapest capable tier, and execute with automatic failover — all in under 50ms.

01. CLASSIFY

Determines task complexity and context length in <4ms.

02. ARBITRAGE

Routes to the lowest latency & cost provider.

03. FAILOVER

Reroutes automatically if primary returns 503 or 429.

Read full architecture docs

Quickstart

POST /v1/chat/completions
curl https://api.meida.cloud/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer me_live_xxxxxxxxxxxxxxxx" \
  -d '{
    "model": "auto-route",
    "messages": [
      { "role": "user", "content": "Analyze this Solana transaction payload..." }
    ],
    "strategy": "cost-latency-balanced"
  }'

Next Steps

Architecture

Understand the 3-step routing pipeline.

Authentication

API key formats and security practices.

Chat Completions

Full endpoint reference with request/response schema.

Routing Strategies

Configure cost vs. latency vs. quality tradeoffs.

Genesis Developer Access

Join the Verified Pilot Queue

Get early API key provisioning and devnet allocation for high-throughput routing.

Back to HomepageArchitecture