long-running low-stakes bulk analysis: background extraction, enrichment and summarization, quality-gated with automatic premium fallback
high-volume execution: bulk drafting and classification at lowest cost
software engineering: writing, editing, and debugging code that must run
documentation and markdown drafting: clear structured prose at low cost
conversation: fast, warm, low-latency dialogue
research: web-grounded investigation, source finding and synthesis at moderate cost
research, analysis, and verification: rigorous reasoning at moderate cost
orchestration and hard reasoning: decomposition, judgment, synthesis of complex work
Last capability watch
No watch has run yet. It fires weekly (and on boot if stale), researching the best and cheapest model per capability across providers. Run one now with the button above.