Model comparison
Claude 3.5 Haiku vs GPT-5 Mini
GPT-5 Mini is the stronger model overall, scoring 41.8 to 29.2 on the Noometry Index.
Last verified . 35 shared benchmarks.
Summary
- They share 35 benchmarks with published results for both. Claude 3.5 Haiku scores higher in 0 categories and GPT-5 Mini in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5 Mini leads 46.7 to 14.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 4.3% for Claude 3.5 Haiku and 86.7% for GPT-5 Mini.
Side by side
| Claude 3.5 Haiku | GPT-5 Mini | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 29.2 | 41.8 |
| Released | 2024-10-22 | 2025-08-07 |
| Weights | Proprietary | Proprietary |
| Context window | — | 400K |
| Max output | — | 128K |
| Input $ / M tokens | — | $0.25 |
| Output $ / M tokens | — | $2 |
| Results tracked | 49 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 Mini leads
Claude 3.5 Haiku: 32.9 (#265), GPT-5 Mini: 40.1 (#146)
| Benchmark | Claude 3.5 Haiku | GPT-5 Mini |
|---|---|---|
| SciCode | 27.4% | 39.2% |
| WeirdML | 30.7% | 52.7% |
| LMArena Coding | 1286 | 1406 |
| SWE-bench Verified | — | 64.7% |
| SWE-bench Verified (bash only) | — | 59.8% |
| Aider Polyglot | 28% | — |
| SWE-bench Multilingual | — | 39.7% |
| BigCodeBench Instruct | 46.1% | — |
| LiveBench Coding | 51.4% | — |
| BigCodeBench Complete | 59% | — |
| CadEval | 32% | — |
| ALE-Bench | — | 799.77 |
| AlgoTune | — | 1.38 |
Agentic & Tool Use GPT-5 Mini leads
Claude 3.5 Haiku: 28.0 (#95), GPT-5 Mini: 31.1 (#70)
| Benchmark | Claude 3.5 Haiku | GPT-5 Mini |
|---|---|---|
| Terminal-Bench | — | 34.8% |
| Berkeley Function Calling Leaderboard | — | 55.5% |
| BALROG | 19.3% | — |
| Vending-Bench 2 | — | -31.18 |
Reasoning GPT-5 Mini leads
Claude 3.5 Haiku: 17.7 (#290), GPT-5 Mini: 23.9 (#168)
| Benchmark | Claude 3.5 Haiku | GPT-5 Mini |
|---|---|---|
| CritPt | 0% | 0% |
| LMArena Hard Prompts | 1251 | 1380 |
| DTBench | 56.7% | 80.5% |
| Epoch Capabilities Index | 127.15 | 145.52 |
| ARC-AGI-2 | — | 4.4% |
| Kagi LLM Benchmark | — | 70.3% |
| ARC-AGI-1 | — | 54.3% |
| Chess Puzzles | — | 30% |
| EnigmaEval | — | 8.2% |
| LiveBench Reasoning | 28.1% | — |
| Mystery Game Puzzles | — | 10% |
| LiveBench Data Analysis | 48.5% | — |
| LMCA | — | 34.2% |
| ForecastBench | — | 61 |
| LiveBench | 43.5% | — |
Math GPT-5 Mini leads
Claude 3.5 Haiku: 14.7 (#300), GPT-5 Mini: 46.7 (#69)
| Benchmark | Claude 3.5 Haiku | GPT-5 Mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 4.3% | 86.7% |
| Omni-MATH | 22.4% | 72.2% |
| LMArena Math | 1244 | 1378 |
| MATH Level 5 | 46.4% | 97.8% |
| FrontierMath (Feb 2025 set) | 0.3% | 27.2% |
| FrontierMath (Tiers 1-3) | — | 46.7% |
| FrontierMath Tier 4 | — | 12.2% |
| ProofBench | — | 9% |
| LiveBench Math | 35.5% | — |
| FrontierMath Tier 4 (v1) | — | 6.3% |
Knowledge GPT-5 Mini leads
Claude 3.5 Haiku: 18.7 (#281), GPT-5 Mini: 45.6 (#86)
| Benchmark | Claude 3.5 Haiku | GPT-5 Mini |
|---|---|---|
| GPQA Diamond | 38.1% | 75% |
| MMLU-Pro | 60.5% | 83.5% |
| Confabulations | 36.7% | 13.3% |
| GPQA (HELM) | 36.3% | 75.6% |
| LMArena Expert | 1208 | 1379 |
| Humanity's Last Exam | — | 19.4% |
| SimpleQA Verified | — | 21.6% |
| Vectara Hallucination Rate | — | 12.9% |
| MMLU | 74.3% | — |
Multimodal GPT-5 Mini leads
Claude 3.5 Haiku: 26.8 (#117), GPT-5 Mini: 35.6 (#85)
| Benchmark | Claude 3.5 Haiku | GPT-5 Mini |
|---|---|---|
| LMArena Vision | 1092 | 1202 |
| GeoBench | 34% | — |
| VPCT | — | 40.2% |
Multilingual GPT-5 Mini leads
Claude 3.5 Haiku: 40.0 (#218), GPT-5 Mini: 48.9 (#137)
| Benchmark | Claude 3.5 Haiku | GPT-5 Mini |
|---|---|---|
| LMArena Non-English | 1238 | 1363 |
| LMArena Chinese | 1229 | 1385 |
| LMArena French | 1264 | 1386 |
| LMArena German | 1237 | 1366 |
| LMArena Japanese | 1175 | 1341 |
| LMArena Korean | 1173 | 1308 |
| LMArena Russian | 1253 | 1362 |
| LMArena Spanish | 1261 | 1355 |
Instruction Following GPT-5 Mini leads
Claude 3.5 Haiku: 62.9 (#234), GPT-5 Mini: 76.2 (#46)
| Benchmark | Claude 3.5 Haiku | GPT-5 Mini |
|---|---|---|
| IFEval | 79.2% | 92.7% |
| LMArena Instruction Following | 1241 | 1357 |
| LiveBench Instruction Following | 61.9% | — |
Long Context GPT-5 Mini leads
Claude 3.5 Haiku: 38.3 (#200), GPT-5 Mini: 41.9 (#132)
| Benchmark | Claude 3.5 Haiku | GPT-5 Mini |
|---|---|---|
| LMArena Longer Query | 1261 | 1355 |
| Fiction.LiveBench | — | 69.4% |
Writing & Preference GPT-5 Mini leads
Claude 3.5 Haiku: 42.7 (#234), GPT-5 Mini: 55.2 (#148)
| Benchmark | Claude 3.5 Haiku | GPT-5 Mini |
|---|---|---|
| LMArena Text | 1255 | 1373 |
| LMArena Creative Writing | 1233 | 1325 |
| Short-Story Creative Writing | 73.5% | 83.1% |
| EQ-Bench Creative Writing | 1146 | 1313 |
| WildBench | 76% | 85.5% |
| LMArena Multi-Turn | 1265 | 1363 |
| LiveBench Language | 35.4% | — |
Frequently asked questions
Is Claude 3.5 Haiku better than GPT-5 Mini?
GPT-5 Mini is the stronger model overall, scoring 41.8 to 29.2 on the Noometry Index.
Is Claude 3.5 Haiku or GPT-5 Mini better for coding?
GPT-5 Mini scores higher on coding benchmarks: 40.1 versus 32.9 in the Noometry coding category.
How many benchmarks do Claude 3.5 Haiku and GPT-5 Mini share?
35 benchmarks have published results for both models. Claude 3.5 Haiku has 49 scored results on Noometry and GPT-5 Mini has 60.