Model comparison
Claude 3 Haiku vs Claude Opus 4
Claude Opus 4 is the stronger model overall, scoring 43.1 to 25.9 on the Noometry Index.
Last verified . 29 shared benchmarks.
Summary
- They share 29 benchmarks with published results for both. Claude 3 Haiku scores higher in 0 categories and Claude Opus 4 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Claude Opus 4 leads 42.0 to 9.8.
- The biggest single-benchmark swing is MATH Level 5: 14.9% for Claude 3 Haiku and 85% for Claude Opus 4.
Side by side
| Claude 3 Haiku | Claude Opus 4 | |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Noometry Index | 25.9 | 43.1 |
| Released | 2024-03-07 | 2025-05-22 |
| Weights | Proprietary | Proprietary |
| Context window | — | 200K |
| Max output | — | 32K |
| Input $ / M tokens | — | $15 |
| Output $ / M tokens | — | $75 |
| Results tracked | 37 | 56 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Opus 4 leads
Claude 3 Haiku: 26.4 (#325), Claude Opus 4: 47.2 (#62)
| Benchmark | Claude 3 Haiku | Claude Opus 4 |
|---|---|---|
| WeirdML | 9.8% | 43.7% |
| LMArena Coding | 1199 | 1442 |
| SWE-bench Verified | — | 70.7% |
| SWE-bench Verified (bash only) | — | 67.6% |
| Aider Polyglot | — | 72% |
| GSO | — | 6.9% |
| BigCodeBench Instruct | 39.4% | — |
| BigCodeBench Complete | 50.1% | — |
| CadEval | 12% | — |
| AlgoTune | — | 1.33 |
| HumanEval+ | 68.9% | — |
| MBPP+ | 68.8% | — |
Agentic & Tool Use Not comparable
Claude 3 Haiku: —, Claude Opus 4: 34.8 (#42)
| Benchmark | Claude 3 Haiku | Claude Opus 4 |
|---|---|---|
| Cybench | — | 38% |
| DeepResearch Bench | — | 46.8% |
| LMArena Search | — | 1127 |
| METR Time Horizons | — | 63.9% |
Reasoning Claude Opus 4 leads
Claude 3 Haiku: 16.3 (#307), Claude Opus 4: 27.3 (#121)
| Benchmark | Claude 3 Haiku | Claude Opus 4 |
|---|---|---|
| Kagi LLM Benchmark | 34.2% | 74.3% |
| LMArena Hard Prompts | 1174 | 1399 |
| DTBench | 50.1% | 81.6% |
| LMCA | 8.8% | 37.4% |
| Epoch Capabilities Index | 118.35 | 142.67 |
| ForecastBench | 53.2 | 61.1 |
| ARC-AGI-2 | — | 8.6% |
| SimpleBench | — | 58.8% |
| ARC-AGI-1 | — | 35.7% |
| CritPt | — | 0.3% |
| EnigmaEval | — | 5.6% |
| WinoGrande | 74.2% | — |
Math Claude Opus 4 leads
Claude 3 Haiku: 9.8 (#319), Claude Opus 4: 42.0 (#86)
| Benchmark | Claude 3 Haiku | Claude Opus 4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.8% | 64.4% |
| LMArena Math | 1188 | 1390 |
| MATH Level 5 | 14.9% | 85% |
| Omni-MATH | — | 61.6% |
| FrontierMath (Feb 2025 set) | — | 4.5% |
| FrontierMath Tier 4 (v1) | — | 4.2% |
Knowledge Claude Opus 4 leads
Claude 3 Haiku: 17.3 (#285), Claude Opus 4: 44.0 (#88)
| Benchmark | Claude 3 Haiku | Claude Opus 4 |
|---|---|---|
| GPQA Diamond | 36.3% | 76.3% |
| Confabulations | 34.2% | 15.9% |
| LMArena Expert | 1148 | 1386 |
| Humanity's Last Exam | — | 10.7% |
| MMLU-Pro | — | 87.5% |
| Vectara Hallucination Rate | — | 12% |
| GPQA (HELM) | — | 70.8% |
| MMLU | 73.8% | — |
Multimodal Claude Opus 4 leads
Claude 3 Haiku: 23.6 (#128), Claude Opus 4: 31.5 (#106)
| Benchmark | Claude 3 Haiku | Claude Opus 4 |
|---|---|---|
| LMArena Vision | 950 | 1192 |
| GeoBench | — | 49% |
| VPCT | — | 38% |
| ScienceQA | 72% | — |
Multilingual Claude Opus 4 leads
Claude 3 Haiku: 36.0 (#243), Claude Opus 4: 48.8 (#138)
| Benchmark | Claude 3 Haiku | Claude Opus 4 |
|---|---|---|
| LMArena Non-English | 1178 | 1362 |
| LMArena Chinese | 1155 | 1386 |
| LMArena French | 1195 | 1372 |
| LMArena German | 1174 | 1391 |
| LMArena Japanese | 1102 | 1331 |
| LMArena Korean | 1109 | 1321 |
| LMArena Russian | 1204 | 1392 |
| LMArena Spanish | 1166 | 1389 |
Instruction Following Claude Opus 4 leads
Claude 3 Haiku: 61.3 (#247), Claude Opus 4: 77.1 (#28)
| Benchmark | Claude 3 Haiku | Claude Opus 4 |
|---|---|---|
| LMArena Instruction Following | 1173 | 1406 |
| IFEval | — | 91.8% |
Long Context Claude Opus 4 leads
Claude 3 Haiku: 36.1 (#237), Claude Opus 4: 39.6 (#172)
| Benchmark | Claude 3 Haiku | Claude Opus 4 |
|---|---|---|
| LMArena Longer Query | 1190 | 1422 |
| Fiction.LiveBench | — | 61.1% |
Writing & Preference Claude Opus 4 leads
Claude 3 Haiku: 29.7 (#291), Claude Opus 4: 61.2 (#89)
| Benchmark | Claude 3 Haiku | Claude Opus 4 |
|---|---|---|
| LMArena Text | 1195 | 1377 |
| LMArena Creative Writing | 1157 | 1387 |
| EQ-Bench Creative Writing | 717 | 1580 |
| LMArena Multi-Turn | 1190 | 1396 |
| Short-Story Creative Writing | — | 83.6% |
| WildBench | — | 85.2% |
Frequently asked questions
Is Claude 3 Haiku better than Claude Opus 4?
Claude Opus 4 is the stronger model overall, scoring 43.1 to 25.9 on the Noometry Index.
Is Claude 3 Haiku or Claude Opus 4 better for coding?
Claude Opus 4 scores higher on coding benchmarks: 47.2 versus 26.4 in the Noometry coding category.
How many benchmarks do Claude 3 Haiku and Claude Opus 4 share?
29 benchmarks have published results for both models. Claude 3 Haiku has 37 scored results on Noometry and Claude Opus 4 has 56.