Willison releases browser-based chat tester for local LLM endpoints

Willison's tool is less notable for its features than for what it assumes: that developers now routinely switch between local inference (LM Studio) and managed APIs (OpenRouter) within the same workflow, and that the OpenAI-compatible endpoint spec has become the de facto interop standard. The gap it fills is not technical but operational.

This is largely disconnected from recent activity in the space we've covered, which makes it worth noting. While major model releases and API announcements dominate headlines, the real consolidation is happening in the tooling layer. When individual developers build utilities to reduce friction between local and cloud inference, it signals that the market has moved past 'which platform do I choose' to 'how do I orchestrate across platforms'. That's a maturity marker. The OpenAI-compatible endpoint becoming a genuine standard (not just lip service) means the competitive moat shifts from API compatibility to model quality, cost, and latency.

If LM Studio, OpenRouter, or another major inference provider ships native support for conversation import/export in this format within six months, it means Willison identified a real workflow need that vendors now see as table stakes. If the tool remains a one-off utility, the friction point was real but not yet large enough to justify vendor investment.