Model comparison
Gemma 4 26B A4B IT vs Llama 3-70B
Gemma 4 26B A4B IT is the stronger model overall, scoring 43.1 to 28.5 on the Noometry Index.
Last verified . 18 shared benchmarks.
Google
MetaSummary
- They share 18 benchmarks with published results for both. Gemma 4 26B A4B IT scores higher in 8 categories and Llama 3-70B in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Gemma 4 26B A4B IT leads 45.7 to 12.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 82.2% for Gemma 4 26B A4B IT and 4.3% for Llama 3-70B.
Side by side
| Gemma 4 26B A4B IT | Llama 3-70B | |
|---|---|---|
| Provider | Meta | |
| Noometry Index | 43.1 | 28.5 |
| Released | 2026-04-02 | 2024-04-18 |
| Weights | Open | Open |
| Context window | 262K | — |
| Max output | 33K | — |
| Input $ / M tokens | $0.0765 | — |
| Output $ / M tokens | $0.26 | — |
| Results tracked | 28 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemma 4 26B A4B IT leads
Gemma 4 26B A4B IT: 38.8 (#156), Llama 3-70B: 35.3 (#215)
| Benchmark | Gemma 4 26B A4B IT | Llama 3-70B |
|---|---|---|
| LMArena Coding | 1446 | 1206 |
| LMArena WebDev | 1359 | — |
| SciCode | 40% | — |
| WeirdML | 35.2% | — |
| BigCodeBench Instruct | — | 43.6% |
| BigCodeBench Complete | — | 54.5% |
| ALE-Bench | 927.17 | — |
| HumanEval+ | — | 72% |
| MBPP+ | — | 69% |
Agentic & Tool Use Not comparable
Gemma 4 26B A4B IT: —, Llama 3-70B: 20.9 (#132)
| Benchmark | Gemma 4 26B A4B IT | Llama 3-70B |
|---|---|---|
| Cybench | — | 5% |
Reasoning Gemma 4 26B A4B IT leads
Gemma 4 26B A4B IT: 21.1 (#205), Llama 3-70B: 17.4 (#284)
| Benchmark | Gemma 4 26B A4B IT | Llama 3-70B |
|---|---|---|
| LMArena Hard Prompts | 1438 | 1195 |
| DTBench | 74.9% | 54.2% |
| Epoch Capabilities Index | 141.85 | 122.91 |
| Kagi LLM Benchmark | — | 35.1% |
| CritPt | 0% | — |
| Chess Puzzles | 6% | — |
| LMCA | 29.7% | — |
| ForecastBench | — | 57.1 |
| WinoGrande | — | 83.5% |
Math Gemma 4 26B A4B IT leads
Gemma 4 26B A4B IT: 45.7 (#61), Llama 3-70B: 12.0 (#299)
| Benchmark | Gemma 4 26B A4B IT | Llama 3-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 82.2% | 4.3% |
| LMArena Math | 1469 | 1218 |
| MATH Level 5 | — | 22.6% |
Knowledge Gemma 4 26B A4B IT leads
Gemma 4 26B A4B IT: 45.6 (#79), Llama 3-70B: 20.8 (#272)
| Benchmark | Gemma 4 26B A4B IT | Llama 3-70B |
|---|---|---|
| GPQA Diamond | 73.2% | 40.6% |
| LMArena Expert | 1447 | 1149 |
| Vectara Hallucination Rate | 5.2% | — |
| MMLU | — | 79.3% |
Multimodal Not comparable
Gemma 4 26B A4B IT: 39.9 (#43), Llama 3-70B: —
| Benchmark | Gemma 4 26B A4B IT | Llama 3-70B |
|---|---|---|
| LMArena Vision | 1260 | — |
Multilingual Gemma 4 26B A4B IT leads
Gemma 4 26B A4B IT: 53.3 (#68), Llama 3-70B: 33.8 (#246)
| Benchmark | Gemma 4 26B A4B IT | Llama 3-70B |
|---|---|---|
| LMArena Non-English | 1422 | 1142 |
| LMArena Chinese | 1495 | 1114 |
| LMArena French | 1459 | 1231 |
| LMArena Russian | 1434 | 1159 |
| LMArena Spanish | 1415 | 1241 |
| LMArena German | — | 1169 |
| LMArena Japanese | — | 1017 |
| LMArena Korean | — | 1017 |
Instruction Following Gemma 4 26B A4B IT leads
Gemma 4 26B A4B IT: 75.0 (#78), Llama 3-70B: 62.8 (#233)
| Benchmark | Gemma 4 26B A4B IT | Llama 3-70B |
|---|---|---|
| LMArena Instruction Following | 1420 | 1194 |
Long Context Gemma 4 26B A4B IT leads
Gemma 4 26B A4B IT: 43.6 (#87), Llama 3-70B: 35.6 (#235)
| Benchmark | Gemma 4 26B A4B IT | Llama 3-70B |
|---|---|---|
| LMArena Longer Query | 1427 | 1174 |
Writing & Preference Gemma 4 26B A4B IT leads
Gemma 4 26B A4B IT: 58.8 (#110), Llama 3-70B: 43.0 (#225)
| Benchmark | Gemma 4 26B A4B IT | Llama 3-70B |
|---|---|---|
| LMArena Text | 1434 | 1221 |
| LMArena Creative Writing | 1401 | 1210 |
| LMArena Multi-Turn | 1442 | 1223 |
| EQ-Bench Creative Writing | 1305 | — |
Frequently asked questions
Is Gemma 4 26B A4B IT better than Llama 3-70B?
Gemma 4 26B A4B IT is the stronger model overall, scoring 43.1 to 28.5 on the Noometry Index.
Is Gemma 4 26B A4B IT or Llama 3-70B better for coding?
Gemma 4 26B A4B IT scores higher on coding benchmarks: 38.8 versus 35.3 in the Noometry coding category.
How many benchmarks do Gemma 4 26B A4B IT and Llama 3-70B share?
18 benchmarks have published results for both models. Gemma 4 26B A4B IT has 28 scored results on Noometry and Llama 3-70B has 31.