Model comparison
DeepSeek-V3.1-Terminus vs Llama 3-8B
DeepSeek-V3.1-Terminus is the stronger model overall, scoring 42.6 to 25.3 on the Noometry Index. Llama 3-8B costs 3.4× less per token, which makes it the better buy when DeepSeek-V3.1-Terminus's lead doesn't matter for your workload.
Last verified . 11 shared benchmarks.
DeepSeek
MetaSummary
- They share 11 benchmarks with published results for both. DeepSeek-V3.1-Terminus scores higher in 7 categories and Llama 3-8B in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek-V3.1-Terminus leads 35.7 to 8.3.
- The biggest single-benchmark swing is DTBench: 81.3% for DeepSeek-V3.1-Terminus and 43.9% for Llama 3-8B.
- Llama 3-8B is cheaper at $0.14 / $0.14 per million input/output tokens, against $0.30 / $1 for DeepSeek-V3.1-Terminus.
- DeepSeek-V3.1-Terminus accepts more context: 164K tokens versus 8K.
Side by side
| DeepSeek-V3.1-Terminus | Llama 3-8B | |
|---|---|---|
| Provider | DeepSeek | Meta |
| Noometry Index | 42.6 | 25.3 |
| Released | 2025-09-22 | 2024-04-18 |
| Weights | Open | Open |
| Context window | 164K | 8K |
| Max output | 66K | 8K |
| Input $ / M tokens | $0.30 | $0.14 |
| Output $ / M tokens | $1 | $0.14 |
| Results tracked | 16 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 41.5 (#114), Llama 3-8B: 30.5 (#285)
| Benchmark | DeepSeek-V3.1-Terminus | Llama 3-8B |
|---|---|---|
| LMArena Coding | 1426 | 1152 |
| SciCode | 40.6% | — |
| BigCodeBench Instruct | — | 31.9% |
| BigCodeBench Complete | — | 36.9% |
| ALE-Bench | 745.17 | — |
| HumanEval+ | — | 56.7% |
| MBPP+ | — | 54.8% |
Reasoning DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 25.7 (#126), Llama 3-8B: 13.8 (#321)
| Benchmark | DeepSeek-V3.1-Terminus | Llama 3-8B |
|---|---|---|
| LMArena Hard Prompts | 1427 | 1133 |
| DTBench | 81.3% | 43.9% |
| Kagi LLM Benchmark | 57.4% | — |
| CritPt | 1.7% | — |
| Chess Puzzles | — | 0% |
| LMCA | 28.6% | — |
| Adversarial NLI | — | 57.3% |
| Epoch Capabilities Index | — | 116.33 |
| ForecastBench | — | 58.6 |
| WinoGrande | — | 75.7% |
Math DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 35.7 (#138), Llama 3-8B: 8.3 (#317)
| Benchmark | DeepSeek-V3.1-Terminus | Llama 3-8B |
|---|---|---|
| LMArena Math | 1402 | 1151 |
| OTIS Mock AIME 2024-2025 | — | 1.9% |
| MATH Level 5 | — | 6.1% |
Knowledge Not comparable
DeepSeek-V3.1-Terminus: —, Llama 3-8B: 7.8 (#303)
| Benchmark | DeepSeek-V3.1-Terminus | Llama 3-8B |
|---|---|---|
| GPQA Diamond | — | 26.1% |
| LMArena Expert | — | 1113 |
| ARC (AI2) Challenge | — | 82.8% |
| MMLU | — | 68.8% |
| OpenBookQA | — | 82.6% |
| TriviaQA | — | 67.7% |
Multilingual DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 52.3 (#87), Llama 3-8B: 31.0 (#256)
| Benchmark | DeepSeek-V3.1-Terminus | Llama 3-8B |
|---|---|---|
| LMArena Non-English | 1407 | 1098 |
| LMArena Russian | 1434 | 1109 |
| LMArena Chinese | — | 1076 |
| LMArena French | — | 1158 |
| LMArena German | — | 1104 |
| LMArena Japanese | — | 967 |
| LMArena Korean | — | 1003 |
| LMArena Spanish | — | 1172 |
Instruction Following DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 74.4 (#96), Llama 3-8B: 58.7 (#255)
| Benchmark | DeepSeek-V3.1-Terminus | Llama 3-8B |
|---|---|---|
| LMArena Instruction Following | 1406 | 1127 |
Long Context DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 43.5 (#88), Llama 3-8B: 34.2 (#246)
| Benchmark | DeepSeek-V3.1-Terminus | Llama 3-8B |
|---|---|---|
| LMArena Longer Query | 1426 | 1128 |
Writing & Preference DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 61.3 (#87), Llama 3-8B: 37.7 (#251)
| Benchmark | DeepSeek-V3.1-Terminus | Llama 3-8B |
|---|---|---|
| LMArena Text | 1420 | 1166 |
| LMArena Creative Writing | 1404 | 1150 |
| LMArena Multi-Turn | 1411 | 1152 |
Frequently asked questions
Is DeepSeek-V3.1-Terminus better than Llama 3-8B?
DeepSeek-V3.1-Terminus is the stronger model overall, scoring 42.6 to 25.3 on the Noometry Index. Llama 3-8B costs 3.4× less per token, which makes it the better buy when DeepSeek-V3.1-Terminus's lead doesn't matter for your workload.
Which is cheaper, DeepSeek-V3.1-Terminus or Llama 3-8B?
Llama 3-8B is cheaper. It lists at $0.14 per million input tokens and $0.14 per million output tokens; DeepSeek-V3.1-Terminus lists at $0.30 and $1.
Is DeepSeek-V3.1-Terminus or Llama 3-8B better for coding?
DeepSeek-V3.1-Terminus scores higher on coding benchmarks: 41.5 versus 30.5 in the Noometry coding category.
Which has the bigger context window?
DeepSeek-V3.1-Terminus does, with 164K tokens against 8K.
How many benchmarks do DeepSeek-V3.1-Terminus and Llama 3-8B share?
11 benchmarks have published results for both models. DeepSeek-V3.1-Terminus has 16 scored results on Noometry and Llama 3-8B has 34.