Cognition's Devin CLI Fusion Cuts Coding Benchmark Costs 39% by Splitting Planner and Executor
DeryaTR_ · x · 2026-09-12
Cognition announced Fusion in Devin CLI, pitched as the most efficient frontier harness for its Fable & Astra models.
- 39% cheaper across coding benchmarks
- Approach: pick a strong model of your choice for planning and a cost-effective model for execution
- Early user feedback: Devin SWE 2.0 alone, without Fusion, already delivers strong results
More from coding & agent
- HABIT-Bench: Best agent memory system scores just 33.61% at inferring user habits — SJTU · 2026-10-28
- Memory Tax study: dedicated agent memory systems often lose to naive RAG — UIUC-CS · 2026-10-28
- Regehr's 'Piano Test': would you trust formally verified code with your life? — moyix · 2026-10-28
- Vibe-coded 'AI 浮生记' game satirizes AI startup life with compute black-market trading — vista8 · 2026-10-28
- Cursor ships Annotate in design mode: agents take visual prompts via on-page markup — soleio · 2026-10-28
- The legendary fake Go generics: Unicode Canadian Syllabics as angle brackets — rickasaurus · 2026-10-28