Model comparison
Claude 3.5 Haiku vs Gemini 3 Pro
Gemini 3 Pro is the stronger model overall, scoring 54.8 to 29.2 on the Noometry Index.
Last verified . 32 shared benchmarks.
GoogleSummary
- They share 32 benchmarks with published results for both. Claude 3.5 Haiku scores higher in 0 categories and Gemini 3 Pro in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Gemini 3 Pro leads 64.4 to 18.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 4.3% for Claude 3.5 Haiku and 91.4% for Gemini 3 Pro.
Side by side
| Claude 3.5 Haiku | Gemini 3 Pro | |
|---|---|---|
| Provider | Anthropic | |
| Noometry Index | 29.2 | 54.8 |
| Released | 2024-10-22 | 2025-11-18 |
| Weights | Proprietary | Proprietary |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 49 | 67 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 3 Pro leads
Claude 3.5 Haiku: 32.9 (#265), Gemini 3 Pro: 51.6 (#39)
| Benchmark | Claude 3.5 Haiku | Gemini 3 Pro |
|---|---|---|
| WeirdML | 30.7% | 69.9% |
| LMArena Coding | 1286 | 1481 |
| SWE-bench Verified | — | 72.9% |
| SWE-bench Verified (bash only) | — | 74.2% |
| Aider Polyglot | 28% | — |
| LMArena WebDev | — | 1440 |
| SWE-bench Multilingual | — | 68.7% |
| SciCode | 27.4% | — |
| GSO | — | 18.6% |
| BigCodeBench Instruct | 46.1% | — |
| LiveBench Coding | 51.4% | — |
| BigCodeBench Complete | 59% | — |
| CadEval | 32% | — |
| ALE-Bench | — | 1,177 |
| AlgoTune | — | 1.83 |
Agentic & Tool Use Gemini 3 Pro leads
Claude 3.5 Haiku: 28.0 (#95), Gemini 3 Pro: 40.6 (#23)
| Benchmark | Claude 3.5 Haiku | Gemini 3 Pro |
|---|---|---|
| BALROG | 19.3% | 58.1% |
| Terminal-Bench | — | 69.4% |
| Berkeley Function Calling Leaderboard | — | 72.5% |
| GDPval | — | 40.3% |
| Remote Labor Index | — | 1.3% |
| τ²-bench Airline | — | 80.5% |
| τ²-bench Banking | — | 18% |
| τ²-bench Retail | — | 75.9% |
| τ²-bench Telecom | — | 91% |
| DeepResearch Bench | — | 46.3% |
| LMArena Search | — | 1207 |
| METR Time Horizons | — | 71% |
| Vending-Bench 2 | — | 5,478 |
Reasoning Gemini 3 Pro leads
Claude 3.5 Haiku: 17.7 (#290), Gemini 3 Pro: 52.5 (#31)
| Benchmark | Claude 3.5 Haiku | Gemini 3 Pro |
|---|---|---|
| CritPt | 0% | 6.9% |
| LMArena Hard Prompts | 1251 | 1480 |
| Epoch Capabilities Index | 127.15 | 152.92 |
| ARC-AGI-2 | — | 31.1% |
| SimpleBench | — | 76.4% |
| Kagi LLM Benchmark | — | 80.1% |
| NYT Connections (extended) | — | 94.4% |
| ARC-AGI-1 | — | 75% |
| Chess Puzzles | — | 31% |
| EnigmaEval | — | 18.2% |
| LiveBench Reasoning | 28.1% | — |
| DTBench | 56.7% | — |
| LiveBench Data Analysis | 48.5% | — |
| ForecastBench | — | 61.2 |
| LiveBench | 43.5% | — |
Math Gemini 3 Pro leads
Claude 3.5 Haiku: 14.7 (#300), Gemini 3 Pro: 49.9 (#59)
| Benchmark | Claude 3.5 Haiku | Gemini 3 Pro |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 4.3% | 91.4% |
| Omni-MATH | 22.4% | 55.5% |
| LMArena Math | 1244 | 1476 |
| FrontierMath (Feb 2025 set) | 0.3% | 37.6% |
| MathArena Final-Answer Competitions | — | 67% |
| ProofBench | — | 20% |
| LiveBench Math | 35.5% | — |
| MATH Level 5 | 46.4% | — |
| FrontierMath Tier 4 (v1) | — | 18.8% |
Knowledge Gemini 3 Pro leads
Claude 3.5 Haiku: 18.7 (#281), Gemini 3 Pro: 64.4 (#16)
| Benchmark | Claude 3.5 Haiku | Gemini 3 Pro |
|---|---|---|
| GPQA Diamond | 38.1% | 92.6% |
| MMLU-Pro | 60.5% | 90.3% |
| GPQA (HELM) | 36.3% | 80.3% |
| LMArena Expert | 1208 | 1475 |
| Humanity's Last Exam | — | 37.5% |
| Confabulations | 36.7% | — |
| Vectara Hallucination Rate | — | 13.6% |
| MMLU | 74.3% | — |
Multimodal Gemini 3 Pro leads
Claude 3.5 Haiku: 26.8 (#117), Gemini 3 Pro: 57.6 (#2)
| Benchmark | Claude 3.5 Haiku | Gemini 3 Pro |
|---|---|---|
| LMArena Vision | 1092 | 1305 |
| GeoBench | 34% | 84% |
| VPCT | — | 91% |
| LMArena Document | — | 1434 |
Multilingual Gemini 3 Pro leads
Claude 3.5 Haiku: 40.0 (#218), Gemini 3 Pro: 56.9 (#16)
| Benchmark | Claude 3.5 Haiku | Gemini 3 Pro |
|---|---|---|
| LMArena Non-English | 1238 | 1474 |
| LMArena Chinese | 1229 | 1523 |
| LMArena French | 1264 | 1492 |
| LMArena German | 1237 | 1515 |
| LMArena Japanese | 1175 | 1510 |
| LMArena Korean | 1173 | 1448 |
| LMArena Russian | 1253 | 1493 |
| LMArena Spanish | 1261 | 1470 |
Instruction Following Gemini 3 Pro leads
Claude 3.5 Haiku: 62.9 (#234), Gemini 3 Pro: 76.3 (#45)
| Benchmark | Claude 3.5 Haiku | Gemini 3 Pro |
|---|---|---|
| IFEval | 79.2% | 87.7% |
| LMArena Instruction Following | 1241 | 1458 |
| LiveBench Instruction Following | 61.9% | — |
Long Context Gemini 3 Pro leads
Claude 3.5 Haiku: 38.3 (#200), Gemini 3 Pro: 44.0 (#79)
| Benchmark | Claude 3.5 Haiku | Gemini 3 Pro |
|---|---|---|
| LMArena Longer Query | 1261 | 1471 |
| CL-bench | — | 15.8% |
Writing & Preference Gemini 3 Pro leads
Claude 3.5 Haiku: 42.7 (#234), Gemini 3 Pro: 66.4 (#35)
| Benchmark | Claude 3.5 Haiku | Gemini 3 Pro |
|---|---|---|
| LMArena Text | 1255 | 1479 |
| LMArena Creative Writing | 1233 | 1482 |
| EQ-Bench Creative Writing | 1146 | 1525 |
| WildBench | 76% | 85.9% |
| LMArena Multi-Turn | 1265 | 1484 |
| Short-Story Creative Writing | 73.5% | — |
| LiveBench Language | 35.4% | — |
Frequently asked questions
Is Claude 3.5 Haiku better than Gemini 3 Pro?
Gemini 3 Pro is the stronger model overall, scoring 54.8 to 29.2 on the Noometry Index.
Is Claude 3.5 Haiku or Gemini 3 Pro better for coding?
Gemini 3 Pro scores higher on coding benchmarks: 51.6 versus 32.9 in the Noometry coding category.
How many benchmarks do Claude 3.5 Haiku and Gemini 3 Pro share?
32 benchmarks have published results for both models. Claude 3.5 Haiku has 49 scored results on Noometry and Gemini 3 Pro has 67.