GPT-6 Astra generates custom running routes from natural language and OSM data
Simon Willison documented GPT-6 Astra's ability to synthesize location data, mapping logic, and geospatial output formats into a practical consumer application. The model processed a natural-language request for running routes, executed a 27-minute reasoning task, and delivered both interactive visualization and machine-readable file formats (GPX, GeoJSON). This signals a meaningful shift in how frontier models handle multi-step spatial reasoning and real-world data integration, moving beyond text generation into domains requiring domain-specific knowledge synthesis and structured output generation.
Modelwire context
ExplainerThe detail worth sitting with is the 27-minute reasoning duration. That is not a conversational exchange; it is closer to a batch job, and it suggests GPT-6 Astra is being used in a mode where latency is traded for depth of planning, which has different cost and UX implications than the instant-response framing most users associate with ChatGPT.
This is largely disconnected from recent activity in our archive, as we have no prior coverage to anchor it to. It belongs to a broader thread in the field around models acquiring reliable tool use and structured output generation, a capability area that has been contested since early function-calling experiments in 2023. Willison's documentation is useful precisely because it is practitioner-led rather than benchmark-led, grounding the claim in a concrete artifact (a usable GPX file) rather than an internal eval score.
Watch whether independent developers can replicate the structured output fidelity (valid, navigable GPX files) consistently across varied geographies in the next 60 days. Isolated success on a familiar route is very different from reliable performance at scale.
This analysis is generated by Modelwire’s editorial layer from our archive and the summary above. It is not a substitute for the original reporting. How we write it.
MentionsOpenAI · GPT-6 Astra · ChatGPT Work · Simon Willison
Modelwire Editorial
This synthesis and analysis was prepared by the Modelwire editorial team. We use advanced language models to read, ground, and connect the day’s most significant AI developments, providing original strategic context that helps practitioners and leaders stay ahead of the frontier.
Modelwire summarizes, we don’t republish. Simon Willison originally reported this story as “Generating running routes with GPT-6 Astra and ChatGPT Work”. The full content lives on simonwillison.net. If you’re a publisher and want a different summarization policy for your work, see our takedown page.