Claude Opus 5.5 vs Llama 3.2 3B (local)
On the 16 use cases both models ran, Claude Opus 5.5 won 16 to 0 (better outcome first, then lower cost).
Pass rate: Claude Opus 5.5 100% (16 runs) vs Llama 3.2 3B (local) 0% (16 runs). Average cost per run: $0.22 vs $0 (local).
| Metric | Claude Opus 5.5 | Llama 3.2 3B (local) |
|---|---|---|
| Runs on shared use cases | 16 | 16 |
| Pass rate | 100% (16/16) | 0% (0/16) |
| Avg cost per run | $0.22 | $0 (local) |
| Avg wall time | 48s | 15s |
| Step-ins per run | 1.69 | 0.00 |
| Use cases won | 16 | 0 |
Use case by use case
- Audit expense claims with per-day caps and unverified exception notes: Claude Opus 5.5 better (pass 100% vs 0%)
- Build a 4-page bakery-cafe website (HTML/CSS): Claude Opus 5.5 better (pass 100% vs 0%)
- Build a landing page from extracted design tokens: Claude Opus 5.5 better (pass 100% vs 0%)
- Build a vendor payout batch when one vendor's amount units are unknown: Claude Opus 5.5 better (pass 100% vs 0%)
- Carry out a customer data-erasure request without erasing a namesake: Claude Opus 5.5 better (pass 100% vs 0%)
- Diagnose why GitHub Actions stopped running: Claude Opus 5.5 better (pass 100% vs 0%)
- Finish a spoken in-car request unattended: Claude Opus 5.5 better (pass 100% vs 0%)
- Implement 5 Python functions from a task list: Claude Opus 5.5 better (pass 100% vs 0%)
- Plan a picking robot's shortest route on a grid with an unconfirmed hazard: Claude Opus 5.5 better (pass 100% vs 0%)
- Rename a data field across a codebase with hidden coupling and a partner-facing export: Claude Opus 5.5 better (pass 100% vs 0%)
- Render per-environment deploy configs when the payment credential is missing: Claude Opus 5.5 better (pass 100% vs 0%)
- Replay a 600-event inventory ledger with an undefined return policy and write-off sign-off: Claude Opus 5.5 better (pass 100% vs 0%)
- Report weekly active users when 'active' and 'week' are still undefined: Claude Opus 5.5 better (pass 100% vs 0%)
- Reverse engineer a security patch for an exploited CVE: Claude Opus 5.5 better (pass 100% vs 0%)
- Roll out catalog price changes with missing categories and sign-off for big increases: Claude Opus 5.5 better (pass 100% vs 0%)
- Schedule a cross-time-zone meeting in a DST-mismatch week and send invites only after sign-off: Claude Opus 5.5 better (pass 100% vs 0%)
Only use cases both models ran are compared; all settings of each model are grouped. How we score.