l1m

l1m is a proxy API that extracts structured JSON data from unstructured text or images using LLMs, eliminating the need for prompt engineering or chat history. It targets developers and data teams in companies of all sizes, offering a schema-first approach where users define their data structure in JSON Schema. The service supports any OpenAI-compatible or Anthropic provider, includes built-in caching, and is open-source with a hosted version available.

Key features

• Schema-first approach with JSON Schema
• Zero prompt engineering required
• Supports OpenAI-compatible and Anthropic providers
• Built-in caching with TTL header
• Open-source, no vendor lock-in
• Hosted version with free tier
• No data retention unless caching used