Clarive Documentation

Clarive is an open-source platform for managing, versioning, and AI-enhancing LLM prompts across teams. Write prompts in a rich editor, organize them with folders and tags, test against multiple AI models, and serve them via a REST API — all from a single workspace.

What you can do with Clarive

  • Write and format prompts — rich text editor with system messages, multi-step prompt chains, and template variables with typed validation.
  • Version and publish — track every change with version history, compare diffs, and publish snapshots served via the API.
  • Experiment with tabs — fork prompt variants into named tabs for A/B testing without affecting the published version.
  • Generate with AI — describe what you need and the AI Wizard writes a complete prompt, scores its quality, and suggests improvements.
  • Test against live models — run prompts in the Playground with streaming responses, batch comparison across models, and AI Judge scoring.
  • Run systematic tests — build a Test Matrix of prompt versions and models, then compare results in a scored grid with downloadable reports.
  • Collaborate in real time — see who's editing, resolve conflicts with keep-mine/keep-theirs/AI-merge, and track activity in the audit log.
  • Integrate via API — fetch and render published prompts with 7 REST endpoints, rate limiting, and official SDKs for C#, Python, and TypeScript.

Browse the docs

© 2026 Clarive. Open source under the MIT License.