Model comparison
DeepSeek-R1-Distill-Llama-70B vs Gemma 2 9B
DeepSeek-R1-Distill-Llama-70B is the stronger model overall, scoring 37.8 to 25.9 on the Noometry Index.
Last verified . 12 shared benchmarks.
DeepSeek
GoogleSummary
- They share 12 benchmarks with published results for both. DeepSeek-R1-Distill-Llama-70B scores higher in 6 categories and Gemma 2 9B in 0 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek-R1-Distill-Llama-70B leads 36.0 to 9.9.
- The biggest single-benchmark swing is MATH Level 5: 89.9% for DeepSeek-R1-Distill-Llama-70B and 21% for Gemma 2 9B.
Side by side
| DeepSeek-R1-Distill-Llama-70B | Gemma 2 9B | |
|---|---|---|
| Provider | DeepSeek | |
| Noometry Index | 37.8 | 25.9 |
| Released | 2025-01-20 | 2024-06-24 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 13 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-R1-Distill-Llama-70B leads
DeepSeek-R1-Distill-Llama-70B: 36.8 (#202), Gemma 2 9B: 29.4 (#304)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Gemma 2 9B |
|---|---|---|
| BigCodeBench Instruct | 35.3% | 34.7% |
| LiveBench Coding | 51.6% | 22.5% |
| BigCodeBench Complete | 49.9% | 40.6% |
| LMArena Coding | — | 1173 |
Reasoning DeepSeek-R1-Distill-Llama-70B leads
DeepSeek-R1-Distill-Llama-70B: 24.9 (#156), Gemma 2 9B: 15.9 (#309)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Gemma 2 9B |
|---|---|---|
| LiveBench Reasoning | 67.6% | 15.2% |
| LiveBench Data Analysis | 55.9% | 36.4% |
| LiveBench | 54.5% | 28.7% |
| Kagi LLM Benchmark | 52.3% | — |
| LMArena Hard Prompts | — | 1171 |
| Epoch Capabilities Index | — | 119.83 |
| PIQA | — | 83.7% |
Math DeepSeek-R1-Distill-Llama-70B leads
DeepSeek-R1-Distill-Llama-70B: 36.0 (#176), Gemma 2 9B: 9.9 (#318)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Gemma 2 9B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 51.4% | 0.6% |
| LiveBench Math | 58.1% | 19.8% |
| MATH Level 5 | 89.9% | 21% |
| LMArena Math | — | 1183 |
| GSM8K | — | 84.9% |
Knowledge DeepSeek-R1-Distill-Llama-70B leads
DeepSeek-R1-Distill-Llama-70B: 30.7 (#225), Gemma 2 9B: 9.7 (#305)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Gemma 2 9B |
|---|---|---|
| GPQA Diamond | 55.7% | 27.5% |
| LMArena Expert | — | 1147 |
| BoolQ | — | 85.7% |
| MMLU | — | 72.1% |
Multilingual Not comparable
DeepSeek-R1-Distill-Llama-70B: —, Gemma 2 9B: 36.6 (#238)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Gemma 2 9B |
|---|---|---|
| LMArena Non-English | — | 1188 |
| LMArena Chinese | — | 1185 |
| LMArena French | — | 1190 |
| LMArena German | — | 1186 |
| LMArena Japanese | — | 1144 |
| LMArena Korean | — | 1137 |
| LMArena Russian | — | 1200 |
| LMArena Spanish | — | 1200 |
Instruction Following DeepSeek-R1-Distill-Llama-70B leads
DeepSeek-R1-Distill-Llama-70B: 68.2 (#190), Gemma 2 9B: 57.6 (#269)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Gemma 2 9B |
|---|---|---|
| LiveBench Instruction Following | 69.9% | 52.6% |
| LMArena Instruction Following | — | 1178 |
Long Context Not comparable
DeepSeek-R1-Distill-Llama-70B: —, Gemma 2 9B: 36.3 (#233)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Gemma 2 9B |
|---|---|---|
| LMArena Longer Query | — | 1197 |
Writing & Preference DeepSeek-R1-Distill-Llama-70B leads
DeepSeek-R1-Distill-Llama-70B: 49.0 (#194), Gemma 2 9B: 32.1 (#281)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Gemma 2 9B |
|---|---|---|
| LiveBench Language | 23.8% | 25.5% |
| LMArena Text | — | 1207 |
| LMArena Creative Writing | — | 1206 |
| EQ-Bench Creative Writing | — | 841 |
| LMArena Multi-Turn | — | 1193 |
Frequently asked questions
Is DeepSeek-R1-Distill-Llama-70B better than Gemma 2 9B?
DeepSeek-R1-Distill-Llama-70B is the stronger model overall, scoring 37.8 to 25.9 on the Noometry Index.
Is DeepSeek-R1-Distill-Llama-70B or Gemma 2 9B better for coding?
DeepSeek-R1-Distill-Llama-70B scores higher on coding benchmarks: 36.8 versus 29.4 in the Noometry coding category.
How many benchmarks do DeepSeek-R1-Distill-Llama-70B and Gemma 2 9B share?
12 benchmarks have published results for both models. DeepSeek-R1-Distill-Llama-70B has 13 scored results on Noometry and Gemma 2 9B has 35.