Model comparison
Gemma 2 2b IT vs GPT-4
Gemma 2 2b IT is the stronger model overall, scoring 32.5 to 28.9 on the Noometry Index.
Last verified . 17 shared benchmarks.
GoogleSummary
- They share 17 benchmarks with published results for both. Gemma 2 2b IT scores higher in 5 categories and GPT-4 in 3 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Gemma 2 2b IT leads 30.1 to 10.1.
- Gemma 2 2b IT has downloadable open weights; the other is API-only.
Side by side
| Gemma 2 2b IT | GPT-4 | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 32.5 | 28.9 |
| Released | — | 2023-03-14 |
| Weights | Open | Proprietary |
| Context window | — | 8K |
| Max output | — | 8K |
| Input $ / M tokens | — | $30 |
| Output $ / M tokens | — | $60 |
| Results tracked | 17 | 38 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Gemma 2 2b IT: 31.8 (#268), GPT-4: 31.2 (#277)
| Benchmark | Gemma 2 2b IT | GPT-4 |
|---|---|---|
| LMArena Coding | 1112 | 1254 |
| WeirdML | — | 12.4% |
| BigCodeBench Instruct | — | 46% |
| BigCodeBench Complete | — | 57.2% |
| HumanEval+ | — | 79.3% |
Agentic & Tool Use Not comparable
Gemma 2 2b IT: —, GPT-4: —
| Benchmark | Gemma 2 2b IT | GPT-4 |
|---|---|---|
| METR Time Horizons | — | 36.1% |
Reasoning Gemma 2 2b IT leads
Gemma 2 2b IT: 20.5 (#219), GPT-4: 17.3 (#285)
| Benchmark | Gemma 2 2b IT | GPT-4 |
|---|---|---|
| LMArena Hard Prompts | 1113 | 1241 |
| Chess Puzzles | — | 4% |
| Mystery Game Puzzles | — | 12% |
| DTBench | — | 62.7% |
| LMCA | — | 17.1% |
| BIG-Bench Hard | — | 75.1% |
| Epoch Capabilities Index | — | 125.88 |
| ForecastBench | — | 57.8 |
| HellaSwag | — | 95.3% |
| WinoGrande | — | 87.5% |
Math Gemma 2 2b IT leads
Gemma 2 2b IT: 30.1 (#212), GPT-4: 10.1 (#304)
| Benchmark | Gemma 2 2b IT | GPT-4 |
|---|---|---|
| LMArena Math | 1135 | 1269 |
| OTIS Mock AIME 2024-2025 | — | 1.1% |
| MATH Level 5 | — | 23% |
| GSM8K | — | 92% |
Knowledge Gemma 2 2b IT leads
Gemma 2 2b IT: 29.3 (#229), GPT-4: 18.4 (#277)
| Benchmark | Gemma 2 2b IT | GPT-4 |
|---|---|---|
| LMArena Expert | 1096 | 1211 |
| GPQA Diamond | — | 35.7% |
| MMLU | — | 86.4% |
| TriviaQA | — | 84.8% |
Multilingual GPT-4 leads
Gemma 2 2b IT: 32.5 (#252), GPT-4: 40.8 (#210)
| Benchmark | Gemma 2 2b IT | GPT-4 |
|---|---|---|
| LMArena Non-English | 1121 | 1246 |
| LMArena Chinese | 1132 | 1242 |
| LMArena French | 1156 | 1282 |
| LMArena German | 1114 | 1251 |
| LMArena Japanese | 1083 | 1209 |
| LMArena Korean | 1055 | 1184 |
| LMArena Russian | 1118 | 1251 |
| LMArena Spanish | 1140 | 1261 |
Instruction Following GPT-4 leads
Gemma 2 2b IT: 58.1 (#258), GPT-4: 65.5 (#217)
| Benchmark | Gemma 2 2b IT | GPT-4 |
|---|---|---|
| LMArena Instruction Following | 1118 | 1241 |
Long Context GPT-4 leads
Gemma 2 2b IT: 34.2 (#245), GPT-4: 37.7 (#207)
| Benchmark | Gemma 2 2b IT | GPT-4 |
|---|---|---|
| LMArena Longer Query | 1130 | 1244 |
Writing & Preference Gemma 2 2b IT leads
Gemma 2 2b IT: 36.6 (#258), GPT-4: 35.0 (#263)
| Benchmark | Gemma 2 2b IT | GPT-4 |
|---|---|---|
| LMArena Text | 1156 | 1263 |
| LMArena Creative Writing | 1147 | 1244 |
| LMArena Multi-Turn | 1118 | 1257 |
| EQ-Bench Creative Writing | — | 752 |
Frequently asked questions
Is Gemma 2 2b IT better than GPT-4?
Gemma 2 2b IT is the stronger model overall, scoring 32.5 to 28.9 on the Noometry Index.
Is Gemma 2 2b IT or GPT-4 better for coding?
They score almost the same on coding (31.8 vs 31.2); test both on your own repository before choosing.
How many benchmarks do Gemma 2 2b IT and GPT-4 share?
17 benchmarks have published results for both models. Gemma 2 2b IT has 17 scored results on Noometry and GPT-4 has 38.