Model comparison
GPT-3.5-turbo vs o3-mini
o3-mini is the stronger model overall, scoring 36.7 to 23.2 on the Noometry Index. GPT-3.5-turbo costs 2.6× less per token, which makes it the better buy when o3-mini's lead doesn't matter for your workload.
Last verified . 28 shared benchmarks.
Summary
- They share 28 benchmarks with published results for both. GPT-3.5-turbo scores higher in 1 category and o3-mini in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where o3-mini leads 38.3 to 10.0.
- The biggest single-benchmark swing is MATH Level 5: 15.9% for GPT-3.5-turbo and 96.5% for o3-mini.
- GPT-3.5-turbo is cheaper at $0.50 / $1.50 per million input/output tokens, against $1.10 / $4.40 for o3-mini.
- o3-mini accepts more context: 200K tokens versus 16K.
Side by side
| GPT-3.5-turbo | o3-mini | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 23.2 | 36.7 |
| Released | 2023-03-01 | 2024-12-20 |
| Weights | Proprietary | Proprietary |
| Context window | 16K | 200K |
| Max output | 4K | 100K |
| Input $ / M tokens | $0.50 | $1.10 |
| Output $ / M tokens | $1.50 | $4.40 |
| Results tracked | 44 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o3-mini leads
GPT-3.5-turbo: 23.9 (#331), o3-mini: 40.8 (#132)
| Benchmark | GPT-3.5-turbo | o3-mini |
|---|---|---|
| WeirdML | 3.5% | 43.7% |
| LMArena Coding | 1136 | 1378 |
| Aider Polyglot | — | 60.4% |
| SciCode | — | 39.8% |
| GSO | — | 1.3% |
| BigCodeBench Instruct | 39.1% | — |
| LiveBench Coding | — | 82.7% |
| BigCodeBench Complete | 50.6% | — |
| CadEval | — | 54% |
| HumanEval+ | 70.7% | — |
| MBPP+ | 69.7% | — |
Agentic & Tool Use Not comparable
GPT-3.5-turbo: —, o3-mini: 29.6 (#84)
| Benchmark | GPT-3.5-turbo | o3-mini |
|---|---|---|
| Cybench | — | 22.5% |
| METR Time Horizons | 21.5% | — |
Reasoning o3-mini leads
GPT-3.5-turbo: 13.8 (#332), o3-mini: 16.3 (#305)
| Benchmark | GPT-3.5-turbo | o3-mini |
|---|---|---|
| Chess Puzzles | 0% | 17% |
| LMArena Hard Prompts | 1108 | 1366 |
| Mystery Game Puzzles | 3% | 7% |
| DTBench | 48.5% | 68.8% |
| LMCA | 9.7% | 19% |
| Epoch Capabilities Index | 118.55 | 140.34 |
| ForecastBench | 50.4 | 59.6 |
| ARC-AGI-2 | — | 3% |
| SimpleBench | — | 22.8% |
| ARC-AGI-1 | — | 34.5% |
| CritPt | — | 0.3% |
| LiveBench Reasoning | — | 89.6% |
| LiveBench Data Analysis | — | 70.6% |
| Adversarial NLI | 58.1% | — |
| BIG-Bench Hard | 61.6% | — |
| CommonsenseQA 2.0 | 57% | — |
| LiveBench | — | 75.9% |
| WinoGrande | 81.6% | — |
Math o3-mini leads
GPT-3.5-turbo: 6.3 (#327), o3-mini: 28.1 (#244)
| Benchmark | GPT-3.5-turbo | o3-mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 0% | 18.6% |
| OTIS Mock AIME 2024-2025 | 2.2% | 76.9% |
| LMArena Math | 1142 | 1396 |
| MATH Level 5 | 15.9% | 96.5% |
| FrontierMath Tier 4 | — | 0% |
| LiveBench Math | — | 77.3% |
| FrontierMath (Feb 2025 set) | — | 12.4% |
| FrontierMath Tier 4 (v1) | — | 4.2% |
| GSM8K | 57.8% | — |
Knowledge o3-mini leads
GPT-3.5-turbo: 10.0 (#303), o3-mini: 38.3 (#146)
| Benchmark | GPT-3.5-turbo | o3-mini |
|---|---|---|
| GPQA Diamond | 28% | 77% |
| LMArena Expert | 1070 | 1364 |
| SimpleQA Verified | — | 15.3% |
| Confabulations | — | 17.9% |
| ARC (AI2) Challenge | 87.4% | — |
| BoolQ | 87% | — |
| MMLU | 71.4% | — |
| OpenBookQA | 86% | — |
| TriviaQA | 85.8% | — |
Multilingual o3-mini leads
GPT-3.5-turbo: 31.5 (#258), o3-mini: 45.7 (#164)
| Benchmark | GPT-3.5-turbo | o3-mini |
|---|---|---|
| LMArena Non-English | 1108 | 1319 |
| LMArena Chinese | 1075 | 1379 |
| LMArena French | 1118 | 1334 |
| LMArena German | 1090 | 1303 |
| LMArena Japanese | 1043 | 1286 |
| LMArena Korean | 1019 | 1314 |
| LMArena Russian | 1123 | 1304 |
| LMArena Spanish | 1121 | 1321 |
Instruction Following o3-mini leads
GPT-3.5-turbo: 57.9 (#262), o3-mini: 75.1 (#72)
| Benchmark | GPT-3.5-turbo | o3-mini |
|---|---|---|
| LMArena Instruction Following | 1119 | 1337 |
| LiveBench Instruction Following | — | 84.4% |
Long Context Too close to call
GPT-3.5-turbo: 34.0 (#254), o3-mini: 33.8 (#256)
| Benchmark | GPT-3.5-turbo | o3-mini |
|---|---|---|
| LMArena Longer Query | 1121 | 1343 |
| Fiction.LiveBench | — | 50% |
Writing & Preference o3-mini leads
GPT-3.5-turbo: 25.3 (#305), o3-mini: 50.3 (#182)
| Benchmark | GPT-3.5-turbo | o3-mini |
|---|---|---|
| LMArena Text | 1125 | 1337 |
| LMArena Creative Writing | 1092 | 1286 |
| LMArena Multi-Turn | 1117 | 1320 |
| Short-Story Creative Writing | — | 61.7% |
| EQ-Bench Creative Writing | 451 | — |
| LiveBench Language | — | 50.7% |
Frequently asked questions
Is GPT-3.5-turbo better than o3-mini?
o3-mini is the stronger model overall, scoring 36.7 to 23.2 on the Noometry Index. GPT-3.5-turbo costs 2.6× less per token, which makes it the better buy when o3-mini's lead doesn't matter for your workload.
Which is cheaper, GPT-3.5-turbo or o3-mini?
GPT-3.5-turbo is cheaper. It lists at $0.50 per million input tokens and $1.50 per million output tokens; o3-mini lists at $1.10 and $4.40.
Is GPT-3.5-turbo or o3-mini better for coding?
o3-mini scores higher on coding benchmarks: 40.8 versus 23.9 in the Noometry coding category.
Which has the bigger context window?
o3-mini does, with 200K tokens against 16K.
How many benchmarks do GPT-3.5-turbo and o3-mini share?
28 benchmarks have published results for both models. GPT-3.5-turbo has 44 scored results on Noometry and o3-mini has 51.