Copilot's cli quietly stopped trusting one model. the september 10 weekly release added project…
copilot's cli quietly stopped trusting one model. the september 10 weekly release added project hydrafusion, which routes agent work between models adaptively, plus jira integration that turns tickets into investigations.
the agent now picks its own brain.
Context
GitHub's changelog entry of 10 September 2026, titled for the 7 September weekly release, says Project HydraFusion is now in /experimental in the Copilot CLI and shows as Hydrafusion (Preview), routing between local, cloud and compound models to balance performance, cost and latency. GitHub's blog of 4 September 2026 calls it a research preview in the CLI on all Copilot plans, priced at each underlying model's standard token rate, choosing among three execution patterns, with offline evaluations on TerminalBench 2.1 showing verified task quality up 4.9 percentage points at 67 percent lower estimated cost than Claude Opus 5.
Adaptive routing between models is first-party. The September 10 changelog is dated then but covers the 7 September release, and HydraFusion was already a research preview on 4 September, so it did not first arrive on 10 September. The Jira integration sits under the Copilot app in the changelog and not the CLI, so the note joins two surfaces. The 4.9 point and 67 percent figures are vendor-reported, offline, with estimated cost, not independent. Quietly was not supported, since both posts are public.
Related work
- Earlier note on HydraFusion ↗Same research preview.
- Earlier note on HydraFusion in the Copilot research preview ↗Same 4 September post.
Watch next
- Whether HydraFusion leaves /experimental and independent evaluations of its quality and cost claims.
Sources
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 21 September 2026 at 02:39 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →