Model comparison
Gemma 7B vs GPT-3.5-turbo
Gemma 7B is the stronger model overall, scoring 29.5 to 23.1 on the Noometry Index.
Last verified . 25 shared benchmarks.
GoogleSummary
- They share 25 benchmarks with published results for both. Gemma 7B scores higher in 5 categories and GPT-3.5-turbo in 3 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Gemma 7B leads 28.7 to 6.0.
- Gemma 7B has downloadable open weights; the other is API-only.
Side by side
| Gemma 7B | GPT-3.5-turbo | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 29.5 | 23.1 |
| Released | 2024-02-21 | 2023-03-01 |
| Weights | Open | Proprietary |
| Context window | — | 16K |
| Max output | — | 4K |
| Input $ / M tokens | — | $0.50 |
| Output $ / M tokens | — | $1.50 |
| Results tracked | 27 | 44 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemma 7B leads
Gemma 7B: 29.9 (#290), GPT-3.5-turbo: 23.5 (#326)
| Benchmark | Gemma 7B | GPT-3.5-turbo |
|---|---|---|
| LMArena Coding | 1048 | 1137 |
| HumanEval+ | 28.7% | 70.7% |
| MBPP+ | 43.4% | 69.7% |
| WeirdML | — | 3.5% |
| BigCodeBench Instruct | — | 39.1% |
| BigCodeBench Complete | — | 50.6% |
Agentic & Tool Use Not comparable
Gemma 7B: —, GPT-3.5-turbo: —
| Benchmark | Gemma 7B | GPT-3.5-turbo |
|---|---|---|
| METR Time Horizons | — | 21.5% |
Reasoning Gemma 7B leads
Gemma 7B: 19.0 (#247), GPT-3.5-turbo: 13.3 (#328)
| Benchmark | Gemma 7B | GPT-3.5-turbo |
|---|---|---|
| LMArena Hard Prompts | 1042 | 1108 |
| Adversarial NLI | 48.7% | 58.1% |
| BIG-Bench Hard | 55.1% | 61.6% |
| Epoch Capabilities Index | 111.74 | 118.48 |
| WinoGrande | 79% | 81.6% |
| Chess Puzzles | — | 0% |
| Mystery Game Puzzles | — | 3% |
| DTBench | — | 48.5% |
| LMCA | — | 9.7% |
| CommonsenseQA 2.0 | — | 57% |
| ForecastBench | — | 50.4 |
| HellaSwag | 82.2% | — |
| PIQA | 81.2% | — |
Math Gemma 7B leads
Gemma 7B: 28.7 (#225), GPT-3.5-turbo: 6.0 (#321)
| Benchmark | Gemma 7B | GPT-3.5-turbo |
|---|---|---|
| LMArena Math | 1066 | 1142 |
| GSM8K | 46.4% | 57.8% |
| FrontierMath (Tiers 1-3) | — | 0% |
| OTIS Mock AIME 2024-2025 | — | 2.2% |
| MATH Level 5 | — | 15.9% |
Knowledge Gemma 7B leads
Gemma 7B: 26.8 (#248), GPT-3.5-turbo: 10.0 (#298)
| Benchmark | Gemma 7B | GPT-3.5-turbo |
|---|---|---|
| LMArena Expert | 1002 | 1070 |
| ARC (AI2) Challenge | 78.3% | 87.4% |
| BoolQ | 83.2% | 87% |
| MMLU | 66.1% | 71.4% |
| OpenBookQA | 78.6% | 86% |
| TriviaQA | 72.3% | 85.8% |
| GPQA Diamond | — | 28% |
Multilingual GPT-3.5-turbo leads
Gemma 7B: 25.3 (#282), GPT-3.5-turbo: 31.6 (#253)
| Benchmark | Gemma 7B | GPT-3.5-turbo |
|---|---|---|
| LMArena Non-English | 999 | 1108 |
| LMArena Chinese | 1035 | 1075 |
| LMArena French | 1024 | 1117 |
| LMArena Russian | 993 | 1123 |
| LMArena German | — | 1090 |
| LMArena Japanese | — | 1043 |
| LMArena Korean | — | 1019 |
| LMArena Spanish | — | 1120 |
Instruction Following GPT-3.5-turbo leads
Gemma 7B: 51.7 (#290), GPT-3.5-turbo: 58.1 (#257)
| Benchmark | Gemma 7B | GPT-3.5-turbo |
|---|---|---|
| LMArena Instruction Following | 1017 | 1119 |
Long Context GPT-3.5-turbo leads
Gemma 7B: 31.1 (#277), GPT-3.5-turbo: 34.0 (#249)
| Benchmark | Gemma 7B | GPT-3.5-turbo |
|---|---|---|
| LMArena Longer Query | 1022 | 1121 |
Writing & Preference Gemma 7B leads
Gemma 7B: 27.2 (#297), GPT-3.5-turbo: 25.4 (#300)
| Benchmark | Gemma 7B | GPT-3.5-turbo |
|---|---|---|
| LMArena Text | 1056 | 1125 |
| LMArena Creative Writing | 1024 | 1092 |
| LMArena Multi-Turn | 963 | 1117 |
| EQ-Bench Creative Writing | — | 451 |
Frequently asked questions
Is Gemma 7B better than GPT-3.5-turbo?
Gemma 7B is the stronger model overall, scoring 29.5 to 23.1 on the Noometry Index.
Is Gemma 7B or GPT-3.5-turbo better for coding?
Gemma 7B scores higher on coding benchmarks: 29.9 versus 23.5 in the Noometry coding category.
How many benchmarks do Gemma 7B and GPT-3.5-turbo share?
25 benchmarks have published results for both models. Gemma 7B has 27 scored results on Noometry and GPT-3.5-turbo has 44.