Sonnet orchestrating local Qwen 3.8 27B cuts costs 2.7x but drops quality from 4/5 to 2/5
GapNew4766 · reddit · 2026-09-29
Testing the popular "cheap local model + expensive cloud model" split on 5 physics-scene builds with one identical prompt: Sonnet 5.5 alone cost $1.83 in 9 min (4/5 scenes), local Qwen 3.8 27B on a 3090 cost $0 in 26 min (1/5, hollow teleporting cubes), and fusion mode — Sonnet writing briefs (5.4K tokens vs 37.9K), Qwen writing all code — cost $0.67 in 85 min (2/5). Reversing roles (Qwen plans, Sonnet codes) hit 4/5 in 14 min but cost more than Sonnet alone since every cloud worker re-reads context. Verdict: the split really cuts the bill 2.7x, paid for in time and quality.
Related event: Cloud-Planning Plus Local-Execution Cuts Costs 2.7x but Takes 9x Longer(3 posts)→
More from coding & agent
- Generic API-catalog MCP server vs one tool per job: a context-bloat experiment — MandeepKaur9407 · 2026-11-16
- Berkeley paper: LLMs answer with stale values even when the update is in context — ombre-ai · 2026-11-16
- Effective HTML: 4.2k-star agent skills turn wireframes into self-contained interactive prototypes — tom_doerr · 2026-11-16
- skills-gate: one gate skill fronts 148 agent skills at just 389 tokens — Distinct-Pie2389 · 2026-11-16
- Open-weight Karpo-Decision-31B tops Decision Index at 73.9, one forward pass per answer — CodeByPoonam · 2026-11-16
- shadcn boosts @uisdk: one-shot UI rebuild that's instantly E2E-testable by AI — shadcn · 2026-11-16