Claude Haiku 5.5 vs Claude Opus 5.5
| Spec | Claude Haiku 5.5 | Claude Opus 5.5 |
|---|---|---|
| Respan Index | 57.2 (#16) | 74.1 (#1) |
| Blended price per 1M | $0.1 | $4 |
| Context window | 1M | 1M |
| Input / output per 1M | $0.05 / $0.25 | $2 / $10 |
| Max output | 128K | 128K |
| Input | text, vision | text, vision |
| Released | 2026-10-07 | 2026-09-22 |
Respan Index by category
Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.
| Category | Claude Haiku 5.5 | Claude Opus 5.5 |
|---|---|---|
| Coding | 55.0 Claude Haiku 5.5 | 68.3 Claude Opus 5.5 |
| Agents | 65.9 Claude Haiku 5.5 | 79.1 Claude Opus 5.5 |
| Reasoning | 65.0 Claude Haiku 5.5 | 83.9 Claude Opus 5.5 |
| Knowledge & language | 23.6 Claude Haiku 5.5 | 70.4 Claude Opus 5.5 |
| Human preference | 55.5 Claude Haiku 5.5 | 61.4 Claude Opus 5.5 |
Benchmarks both report
Claude Opus 5.5 leads on 27 of the 29 benchmarks both report. Higher is better. Each score links to where it was published.
| Benchmark | Claude Haiku 5.5 | Claude Opus 5.5 |
|---|---|---|
| MirrorCode (Epoch AI) | 78.9% | 78.9% |
| SAGE 1.1 (Vals AI) | 89.8% | 91.3% |
| APEX-Agents (Mercor) | 56.1% | 73.5% |
| Chess Puzzles (Epoch AI) | 55% | 67% |
| LMArena Agent (LMArena) | 0.0143 | 0.2094 |
| LMArena Elo | 1425 | 1509 |
| LMArena WebDev (LMArena) | 1589 | 1810 |
| Terminal-Bench 4.0 (Vals AI) | 36.4% | 65.2% |
| Vending-Bench 2 (Andon Labs) | 3602.46 | 9235.25 |
| EBR-bench (Epoch AI) | 43.8% | 71.4% |
| FrontierMath Tier 4 v2 (Epoch AI) | 46.3% | 95% |
| FrontierMath Tiers 1-3 v2 (Epoch AI) | 75.1% | 91.2% |
| FrontierSWE V2 (Proximal Labs) | 43.8% | 62.3% |
| Furniture Assembly (Epoch AI) | 47.5% | 83.3% |
| GPQA Diamond | 89.6% | 90.6% |
| LiveBench Agentic Coding (LiveBench) | 57.2% | 71.7% |
| LiveBench Coding (LiveBench) | 77.9% | 89.3% |
| LiveBench Data Analysis (LiveBench) | 72.8% | 80.3% |
| LiveBench Instruction Following (LiveBench) | 66.5% | 67% |
| LiveBench Language (LiveBench) | 63.5% | 86.3% |
| LiveBench Mathematics (LiveBench) | 92.8% | 97.1% |
| LiveBench Reasoning (LiveBench) | 84.6% | 92.2% |
| Mystery Game Puzzles (Epoch AI) | 30% | 71% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 98.9% | 100% |
| SAGE (Vals AI) | 45.8% | 45.8% |
| SimpleQA Verified | 23.8% | 72.2% |
| FrontierCode v1.1 (Main) | 46.4% | 54.4% |
| Humanity's Last Exam (with tools) | 57.4% | 67.7% |
| Terminal-Bench 4.0 | 39.2% | 66.4% |
FAQ
Is Claude Haiku 5.5 better than Claude Opus 5.5?
Claude Opus 5.5 scores 74.1 on the Respan Index (#1), ahead of Claude Haiku 5.5 at 57.2 (#16). Claude Opus 5.5 leads on 27 of the 29 benchmarks both report.
Which is cheaper, Claude Haiku 5.5 or Claude Opus 5.5?
Claude Haiku 5.5 lists at $0.05 input and $0.25 output per 1M tokens; Claude Opus 5.5 at $2 and $10. Claude Haiku 5.5 is 40x cheaper per token at list price.
Which has the larger context window?
Both accept 1M tokens of context.