Blind test: Sonnet 5.5 matches Opus 5.5 on a from-scratch Rust decompressor at 1/4 the price
_Duex · reddit · 2026-09-29
The author benchmarked Sonnet 5.5 vs Opus 5.5 (plus GPT-6 Sol/Luna) on the same self-contained task: write a dependency-free, unsafe-free DEFLATE/zlib decompressor in Rust, graded blind and fully automated against zlib with 4055 hidden tests (real streams, hand-built edge cases, 4000 corrupted inputs, a speed test, a zip bomb).
- Opus 5.5: all passed, 507 MB/s, 10m18s, $2.08
- Sonnet 5.5: all passed, 447 MB/s, 3m41s, $0.52
- GPT-6 Sol: all passed, 301 MB/s, 6m37s, $0.27
- GPT-6 Luna: failed on normal zlib (one typo), $0.007
Opus earned its price on polish—zlib-style two-level Huffman tables, buffer discipline, and a self-written differential tester. For well-defined tasks, Sonnet 5.5 at medium looks like the better deal; Opus may still lead on open-ended debugging. Single task, single run per model.
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