Github's new hydrafusion routes a single coding task across multiple models instead of betting on…
github's new hydrafusion routes a single coding task across multiple models instead of betting on one, and github's own tests claim 36-67% lower cost than opus 5. the "which model do you use" question is about to sound quaint.
enterprises won't buy models anymore, they'll buy the router.
Context
GitHub's post of 4 September 2026 describes Project HydraFusion as a research preview in GitHub Copilot, on all plans through /experimental in Copilot CLI, billed at each underlying model's standard rate. It picks one of three patterns per request, Single, Cascade or Critique, across models from multiple providers. In GitHub's own offline evaluation with fixed policies, medium reasoning and Opus 5 as the baseline, it reports 67% lower estimated cost with 4.9 points higher on TerminalBench 2.1, 36% lower with 1.5 points lower on DeepSWE, and 65% lower with 0.1 points lower on CheckpointBench, GitHub's internal benchmark.
These are vendor-run offline evaluations with estimated costs, specific to the benchmark revisions, configurations, model pool and pricing assumptions, and not independent. The routing policies were tuned across the same three benchmarks, and GitHub says TerminalBench 2.1 is relatively saturated. The 36 to 67% range leaves out that quality is lower on two of the three. Routing across models applies to Cascade and Critique, and Single uses one model. The post is dated 4 September, and the later availability status was not re-read here, so it is not presented as current. Enterprises buying the router is the author's forecast.
Related work
- Earlier note on HydraFusion ↗Same GitHub post.
Watch next
- Independent replication and later availability changes.
Sources
- Project HydraFusion (GitHub, 4 Sep 2026)github.blog
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 19 September 2026 at 05:16 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 →