GitHub HydraFusion adds Max tier, model pool now includes GPT-6 family and Claude Opus 5.5
lee_stott · x · 2026-10-15
GitHub updated HydraFusion, its research-preview feature that orchestrates multiple models into a single response so users don't pick models manually:
- Updated model pool: HydraFusion and the new HydraFusion Max now draw from Claude Opus 5.5, GPT-6 Astra, GPT-6.1 Sol, and GPT-6 Luna, selected per orchestration workflow.
- Per-stage reasoning effort: instead of default effort, initial drafts use medium effort while review uses high effort, balancing efficiency and depth.
- Workflow upgrades: Cascade gains a judging step that requests repairs only when needed; Critique relies more on tools inspecting the actual workspace and adds parallel critic models.
- Max tier: same architecture tuned for frontier capability on demanding tasks with higher reasoning effort.
Both tiers work across VS Code, the Copilot app, and Copilot CLI, and remain in research preview.
More from coding & agent
- SWEeper-Bench: Best of 15 Frontier Agents Solves Only 59% of Interactive Bug-Finding Tasks — zlab-princeton · 2026-10-15
- User hands Google login and SMS 2FA to AI agent, reigniting credential-handoff debate — jonstephens85 · 2026-10-15
- Grok Bot writes its own Apple Desk skill to control Mac apps, Musk amplifies — elonmusk · 2026-10-15
- Thorsten Ball: latest models make context-management tricks like /btw and forking obsolete — andreisavu · 2026-10-15
- User says Grok Bot finished 10 days of work in under 2 hours on a $30 plan — elonmusk · 2026-10-15
- PersonalAgentEval launches: 7 personal agents tested on real 13-month histories — HennyJieCC · 2026-10-15