What it does
api100 routes requests to models from providers such as OpenAI, Anthropic, Google, and DeepSeek. Applications use an official OpenAI library in TypeScript, Python, Go, or cURL. They point the base URL to https://api.apihundred.com/v1 and authenticate with an api100 API key. Requests go to the chat completions endpoint with an OpenAI-compatible payload. The router checks the API key and credit balance with PostgreSQL triggers. It health-checks routes and retries on alternate nodes when it detects upstream 429 rate limits or timeouts. Tokens stream back to the client with Server-Sent Events. The credit ledger is deducted when the response completes.
Capabilities
- The same endpoint serves model names such as claude-3-5-sonnet, deepseek-r1, gpt-4o, and gemini-1-5-pro.
- One prepaid balance powers Claude 3.5 Sonnet, GPT-4o, DeepSeek R1, and Gemini 1.5 Pro.
- The telemetry ledger shows latency, prompt tokens, completion tokens, and exact dollar cost per API key, application route, or team member.
- The gateway handles retries and endpoint re-routing during upstream provider downtime without requiring application restarts.
- api100 says credits never expire and rollover automatically.
When to use it
- Use it when you want to stop managing fragmented accounts across OpenAI, Anthropic, Google, and DeepSeek.
- Use it when you want to migrate an existing OpenAI library integration by changing the base URL.
- Use it when you want one gateway to retry and re-route requests during Anthropic or OpenAI downtime.
