ThoughtDAG
An editable context graph for LLM conversations
Last verified:
What is ThoughtDAG?
ThoughtDAG is a visual graph-based interface that makes LLM context explicit and editable. Users can see exactly which conversation history and sources will be sent to the model, then branch conversations, prune irrelevant paths, and merge reasoning threads—transforming hidden context selection into a visible, auditable protocol.
ThoughtDAG pricing
Pricing model: Freemium
ThoughtDAG pros
- Visible context inspection with token preview before sending to the model
- Editable context graph with branch, prune, and merge operations—no overwriting of prior reasoning paths
- Reproducible results: delete one edge and regenerate to verify how context changes affect answers
- Desktop application for Linux, macOS, and Windows with bundled local engine
ThoughtDAG cons
- Manual graph curation required—users must actively manage context structure
- Windows builds are not code-signed (triggers SmartScreen warning)
- No documented pricing or tier information; integration options with external LLM services unclear
Frequently asked questions about ThoughtDAG
What does ThoughtDAG do?
It turns LLM conversation context into an editable graph so you can see and control exactly what the model receives, rather than having context selection be a hidden decision.
Can I run ThoughtDAG offline?
Yes. Desktop versions for Linux, macOS (Apple Silicon and Intel), and Windows include a bundled local engine.
What happens when I delete an edge from the context graph?
That branch is excluded from the next request. You can regenerate the same prompt and observe how the answer changes with the modified context.