Model comparison
DeepSeek-V3.1 vs Llama 3-8B
DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 25.5 on the Noometry Index.
Last verified . 20 shared benchmarks.
DeepSeek
MetaSummary
- They share 20 benchmarks with published results for both. DeepSeek-V3.1 scores higher in 8 categories and Llama 3-8B in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek-V3.1 leads 43.7 to 7.8.
- The biggest single-benchmark swing is DTBench: 82.7% for DeepSeek-V3.1 and 43.9% for Llama 3-8B.
Side by side
| DeepSeek-V3.1 | Llama 3-8B | |
|---|---|---|
| Provider | DeepSeek | Meta |
| Noometry Index | 42.8 | 25.5 |
| Released | 2025-08-21 | 2024-04-18 |
| Weights | Open | Open |
| Context window | 164K | — |
| Max output | 8K | — |
| Input $ / M tokens | $0.25 | — |
| Output $ / M tokens | $0.95 | — |
| Results tracked | 27 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.1 leads
DeepSeek-V3.1: 40.3 (#144), Llama 3-8B: 31.0 (#289)
| Benchmark | DeepSeek-V3.1 | Llama 3-8B |
|---|---|---|
| LMArena Coding | 1417 | 1152 |
| WeirdML | 38.4% | — |
| BigCodeBench Instruct | — | 31.9% |
| BigCodeBench Complete | — | 36.9% |
| HumanEval+ | — | 56.7% |
| MBPP+ | — | 54.8% |
Reasoning DeepSeek-V3.1 leads
DeepSeek-V3.1: 27.9 (#110), Llama 3-8B: 14.3 (#326)
| Benchmark | DeepSeek-V3.1 | Llama 3-8B |
|---|---|---|
| LMArena Hard Prompts | 1417 | 1133 |
| DTBench | 82.7% | 43.9% |
| Epoch Capabilities Index | 139.92 | 116.45 |
| ForecastBench | 58 | 58.6 |
| SimpleBench | 40% | — |
| Kagi LLM Benchmark | 53.2% | — |
| Chess Puzzles | — | 0% |
| LMCA | 24.3% | — |
| Adversarial NLI | — | 57.3% |
| WinoGrande | — | 75.7% |
Math DeepSeek-V3.1 leads
DeepSeek-V3.1: 38.9 (#122), Llama 3-8B: 8.8 (#323)
| Benchmark | DeepSeek-V3.1 | Llama 3-8B |
|---|---|---|
| LMArena Math | 1420 | 1151 |
| OTIS Mock AIME 2024-2025 | — | 1.9% |
| MATH Level 5 | — | 6.1% |
Knowledge DeepSeek-V3.1 leads
DeepSeek-V3.1: 43.7 (#90), Llama 3-8B: 7.8 (#308)
| Benchmark | DeepSeek-V3.1 | Llama 3-8B |
|---|---|---|
| LMArena Expert | 1405 | 1113 |
| GPQA Diamond | — | 26.1% |
| Vectara Hallucination Rate | 5.5% | — |
| ARC (AI2) Challenge | — | 82.8% |
| MMLU | — | 68.8% |
| OpenBookQA | — | 82.6% |
| TriviaQA | — | 67.7% |
Multilingual DeepSeek-V3.1 leads
DeepSeek-V3.1: 51.6 (#106), Llama 3-8B: 30.8 (#261)
| Benchmark | DeepSeek-V3.1 | Llama 3-8B |
|---|---|---|
| LMArena Non-English | 1400 | 1098 |
| LMArena Chinese | 1469 | 1076 |
| LMArena French | 1447 | 1159 |
| LMArena German | 1411 | 1104 |
| LMArena Japanese | 1378 | 967 |
| LMArena Korean | 1337 | 1004 |
| LMArena Russian | 1405 | 1109 |
| LMArena Spanish | 1431 | 1173 |
Instruction Following DeepSeek-V3.1 leads
DeepSeek-V3.1: 73.9 (#110), Llama 3-8B: 58.4 (#260)
| Benchmark | DeepSeek-V3.1 | Llama 3-8B |
|---|---|---|
| LMArena Instruction Following | 1400 | 1127 |
Long Context DeepSeek-V3.1 leads
DeepSeek-V3.1: 36.3 (#232), Llama 3-8B: 34.2 (#251)
| Benchmark | DeepSeek-V3.1 | Llama 3-8B |
|---|---|---|
| LMArena Longer Query | 1422 | 1128 |
| Fiction.LiveBench | 52.8% | — |
Writing & Preference DeepSeek-V3.1 leads
DeepSeek-V3.1: 60.3 (#98), Llama 3-8B: 37.5 (#256)
| Benchmark | DeepSeek-V3.1 | Llama 3-8B |
|---|---|---|
| LMArena Text | 1420 | 1166 |
| LMArena Creative Writing | 1401 | 1150 |
| LMArena Multi-Turn | 1408 | 1152 |
| EQ-Bench Creative Writing | 1436 | — |
Frequently asked questions
Is DeepSeek-V3.1 better than Llama 3-8B?
DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 25.5 on the Noometry Index.
Is DeepSeek-V3.1 or Llama 3-8B better for coding?
DeepSeek-V3.1 scores higher on coding benchmarks: 40.3 versus 31.0 in the Noometry coding category.
How many benchmarks do DeepSeek-V3.1 and Llama 3-8B share?
20 benchmarks have published results for both models. DeepSeek-V3.1 has 27 scored results on Noometry and Llama 3-8B has 34.