Model comparison
Claude Haiku 4.5 vs Claude Opus 5.5
Claude Opus 5.5 is the stronger model overall, scoring 68.6 to 39.5 on the Noometry Index. Claude Haiku 4.5 costs 4.0× less per token, which makes it the better buy when Claude Opus 5.5's lead doesn't matter for your workload.
Last verified . 29 shared benchmarks.
Summary
- They share 29 benchmarks with published results for both. Claude Haiku 4.5 scores higher in 0 categories and Claude Opus 5.5 in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Claude Opus 5.5 leads 80.2 to 15.1.
- The biggest single-benchmark swing is ARC-AGI-2: 4% for Claude Haiku 4.5 and 93.3% for Claude Opus 5.5.
- Claude Haiku 4.5 is cheaper at $1 / $5 per million input/output tokens, against $4 / $20 for Claude Opus 5.5.
- Claude Opus 5.5 accepts more context: 1M tokens versus 200K.
Side by side
| Claude Haiku 4.5 | Claude Opus 5.5 | |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Noometry Index | 39.5 | 68.6 |
| Released | 2025-10-15 | 2026-09-22 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 1M |
| Max output | 64K | 128K |
| Input $ / M tokens | $1 | $4 |
| Output $ / M tokens | $5 | $20 |
| Results tracked | 53 | 44 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Opus 5.5 leads
Claude Haiku 4.5: 44.0 (#78), Claude Opus 5.5: 71.9 (#3)
| Benchmark | Claude Haiku 4.5 | Claude Opus 5.5 |
|---|---|---|
| LMArena WebDev | 1330 | 1813 |
| SciCode | 43.3% | 66.9% |
| LMArena Coding | 1453 | 1547 |
| ALE-Bench | 653.48 | 2,147 |
| FrontierCode | — | 54.6% |
| SWE-bench Verified (bash only) | 66.6% | — |
| CursorBench | — | 57.8% |
| SWE-bench Multilingual | 64.7% | — |
| FrontierSWE | — | 62.3% |
| WeirdML | 45.4% | — |
| MirrorCode | — | 77.4% |
Agentic & Tool Use Claude Opus 5.5 leads
Claude Haiku 4.5: 33.6 (#52), Claude Opus 5.5: 45.3 (#15)
| Benchmark | Claude Haiku 4.5 | Claude Opus 5.5 |
|---|---|---|
| Vending-Bench 2 | 458.89 | 9,235 |
| Terminal-Bench | 35.5% | — |
| APEX-Agents | — | 73.5% |
| Berkeley Function Calling Leaderboard | 68.7% | — |
| DeepResearch Bench | 45.5% | — |
| BALROG | 31.2% | — |
| ExploitBench | 13.7% | — |
| GDP.pdf | — | 30.6% |
Reasoning Claude Opus 5.5 leads
Claude Haiku 4.5: 15.1 (#320), Claude Opus 5.5: 80.2 (#3)
| Benchmark | Claude Haiku 4.5 | Claude Opus 5.5 |
|---|---|---|
| ARC-AGI-2 | 4% | 93.3% |
| NYT Connections (extended) | 14.3% | 88.5% |
| ARC-AGI-1 | 47.7% | 98.5% |
| CritPt | 0% | 31.7% |
| LMArena Hard Prompts | 1420 | 1535 |
| DTBench | 73.6% | 98.9% |
| LMCA | 30.9% | 68.2% |
| Epoch Capabilities Index | 142.41 | 167.33 |
| Chess Puzzles | 8% | — |
| EBR-Bench | — | 71.4% |
| Mystery Game Puzzles | — | 71% |
| ForecastBench | 61.4 | — |
Math Claude Opus 5.5 leads
Claude Haiku 4.5: 44.9 (#78), Claude Opus 5.5: 91.8 (#3)
| Benchmark | Claude Haiku 4.5 | Claude Opus 5.5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 66.7% | 100% |
| LMArena Math | 1396 | 1506 |
| FrontierMath (Tiers 1-3) | — | 91.2% |
| FrontierMath Tier 4 | — | 95% |
| ProofBench | — | 100% |
| Omni-MATH | 56.1% | — |
| MATH Level 5 | 96.4% | — |
| FrontierMath (Feb 2025 set) | 5.9% | — |
| FrontierMath Erdős | — | 2.9% |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge Claude Opus 5.5 leads
Claude Haiku 4.5: 37.7 (#153), Claude Opus 5.5: 66.4 (#10)
| Benchmark | Claude Haiku 4.5 | Claude Opus 5.5 |
|---|---|---|
| GPQA Diamond | 71.2% | 90.6% |
| SimpleQA Verified | 13.2% | 72.2% |
| LMArena Expert | 1442 | 1547 |
| MMLU-Pro | 77.7% | — |
| Vectara Hallucination Rate | 9.8% | — |
| GPQA (HELM) | 60.5% | — |
Multimodal Claude Opus 5.5 leads
Claude Haiku 4.5: 26.8 (#118), Claude Opus 5.5: 57.8 (#1)
| Benchmark | Claude Haiku 4.5 | Claude Opus 5.5 |
|---|---|---|
| Blueprint-Bench 2 | 0% | 51.2% |
| LMArena Vision | — | 1321 |
| Furniture Assembly | — | 83.3% |
| LMArena Document | 1420 | — |
Multilingual Claude Opus 5.5 leads
Claude Haiku 4.5: 49.9 (#129), Claude Opus 5.5: 59.1 (#2)
| Benchmark | Claude Haiku 4.5 | Claude Opus 5.5 |
|---|---|---|
| LMArena Non-English | 1377 | 1507 |
| LMArena Chinese | 1417 | 1588 |
| LMArena French | 1408 | 1514 |
| LMArena Russian | 1381 | 1520 |
| LMArena Spanish | 1420 | 1507 |
| LMArena German | 1375 | — |
| LMArena Japanese | 1339 | — |
| LMArena Korean | 1347 | — |
Instruction Following Claude Opus 5.5 leads
Claude Haiku 4.5: 71.4 (#149), Claude Opus 5.5: 80.0 (#3)
| Benchmark | Claude Haiku 4.5 | Claude Opus 5.5 |
|---|---|---|
| LMArena Instruction Following | 1414 | 1537 |
| IFEval | 80.1% | — |
Long Context Claude Opus 5.5 leads
Claude Haiku 4.5: 43.6 (#92), Claude Opus 5.5: 47.1 (#19)
| Benchmark | Claude Haiku 4.5 | Claude Opus 5.5 |
|---|---|---|
| LMArena Longer Query | 1427 | 1532 |
Writing & Preference Claude Opus 5.5 leads
Claude Haiku 4.5: 57.9 (#123), Claude Opus 5.5: 78.2 (#3)
| Benchmark | Claude Haiku 4.5 | Claude Opus 5.5 |
|---|---|---|
| LMArena Text | 1396 | 1515 |
| LMArena Creative Writing | 1372 | 1533 |
| LMArena Multi-Turn | 1409 | 1499 |
| EQ-Bench Creative Writing | — | 2050 |
| WildBench | 83.9% | — |
| EQ-Bench 4 | 1064 | — |
Frequently asked questions
Is Claude Haiku 4.5 better than Claude Opus 5.5?
Claude Opus 5.5 is the stronger model overall, scoring 68.6 to 39.5 on the Noometry Index. Claude Haiku 4.5 costs 4.0× less per token, which makes it the better buy when Claude Opus 5.5's lead doesn't matter for your workload.
Which is cheaper, Claude Haiku 4.5 or Claude Opus 5.5?
Claude Haiku 4.5 is cheaper. It lists at $1 per million input tokens and $5 per million output tokens; Claude Opus 5.5 lists at $4 and $20.
Is Claude Haiku 4.5 or Claude Opus 5.5 better for coding?
Claude Opus 5.5 scores higher on coding benchmarks: 71.9 versus 44.0 in the Noometry coding category.
Which has the bigger context window?
Claude Opus 5.5 does, with 1M tokens against 200K.
How many benchmarks do Claude Haiku 4.5 and Claude Opus 5.5 share?
29 benchmarks have published results for both models. Claude Haiku 4.5 has 53 scored results on Noometry and Claude Opus 5.5 has 44.