Langfuse
93 · HealthyTracing, evaluation and prompt management for LLM applications and agents.
Platforms for building LLM chatbots, agents and RAG workflows. Ranked by health score.
Tracing, evaluation and prompt management for LLM applications and agents.
Terminal coding agent that works with many model providers and can edit, run and test code in your repository.
Observability, evaluations and agent simulations for teams running LLM agents in production.
AI search and chat over your company's documents, with connectors for Slack, Google Drive, Confluence and more.
Studio for building LLM chatbots and agents, with knowledge bases and connectors for web chat, WhatsApp and other channels.
AI coding agent you can run yourself. Give it an issue and it edits code, runs commands and tests in a sandbox, then opens a pull request, with whichever LLM you choose.
Autonomous coding agent for VS Code and the command line that plans changes, edits files and runs commands with your approval.
Framework for customer-facing AI agents that follow explicit guidelines, so their answers stay consistent and on policy.
Visual builder for AI agents and RAG workflows, with a Python API.
Research agent that takes a GitHub issue and tries to fix it automatically with the language model of your choice.
Voice AI platform for building phone agents, with a visual workflow editor, telephony support and your choice of speech and LLM providers.
Proxy-based LLM observability that logs requests, costs and latency with one line of setup.
AI pair programmer in the terminal that edits code in your git repository and commits each change.
Drag-and-drop builder for LLM agents and chatflows.
The code for these is public, but the license is not an OSI open source license, or we have not verified it yet. Check the license notes on each project before relying on it.






Health and license are separate verdicts: a project can be well maintained and still not be open source.
| Project | Health | License | Stars |
|---|---|---|---|
| Langfuse | 93 · Healthy | Free core, paid extrasMIT | 36k |
| opencode | 92 · Healthy | Free to useMIT | 212k |
| LangWatch | 89 · Healthy | Free core, paid extrasApache-2.0 | 4.9k |
| Onyx | 88 · Healthy | Free core, paid extrasMIT | 32k |
| Botpress | 86 · Healthy | Free to useMIT | 15k |
| OpenHands | 85 · Healthy | Free to useMIT | 90k |
| Cline | 83 · Healthy | Free to useApache-2.0 | 70k |
| Parlant | 79 · Stable | Free to useApache-2.0 | 18k |
| Langflow | 77 · Stable | Free to useMIT | 155k |
| SWE-agent | 73 · Stable | Free to useMIT | 21k |
| Dograh | 72 · Stable | Free to useBSD-2-Clause | 5.8k |
| Helicone | 67 · Stable | Free to useApache-2.0 | 6.2k |
| Aider | 48 · Caution | Free to useApache-2.0 | 49k |
| Flowise | Archived | Free core, paid extrasApache-2.0 | 55k |
| Dify | 91 · Healthy | Free with conditionsDify Open Source License | 158k |
| Phoenix | 77 · Stable | Free, no resellingElastic License 2.0 | 12k |
Alternatives to BraintrustAlternatives to ChatbaseAlternatives to ChatfuelAlternatives to DevinAlternatives to DocsBot AIAlternatives to LandbotAlternatives to LangSmithAlternatives to LindyAlternatives to ManyChatAlternatives to Microsoft Copilot StudioAlternatives to Relevance AIAlternatives to Retell AIAlternatives to VapiAlternatives to Voiceflow
Scores last refreshed from GitHub on 9 October 2026. How we score.