Glassbrain

Glassbrain provides an interactive visual interface for debugging AI applications that use LLM services such as OpenAI, Anthropic, and LangChain. It records each call in a trace, presents the full reasoning chain as a clickable tree, and lets users replay any step with modified inputs without redeploying code. The tool highlights the root cause of incorrect outputs, offers concrete fix suggestions generated by a Claude model, and shows side‑by‑side diffs of before and after changes.

The product targets developers who build and maintain AI‑driven workflows and need to diagnose failures quickly. By converting long log files into an explorable visual trace, it reduces the time spent parsing text logs and enables one‑click fixes, allowing teams to resolve bugs in seconds rather than minutes.

Installation is a single npm command followed by an import, and the system integrates with existing stacks via an OpenTelemetry‑compatible endpoint. It supports a range of providers, including OpenAI, Anthropic, LangChain, and LlamaIndex, and can be used with custom instrumentation. Pricing scales with usage, offering a free tier with limited traces and features.