Model comparison
DeepSeek-V2.5 (Sep 2024) vs Llama 4 Scout
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.6 to 27.7 on the Noometry Index.
Last verified . 18 shared benchmarks.
DeepSeek
MetaSummary
- They share 18 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 8 categories and Llama 4 Scout in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek-V2.5 (Sep 2024) leads 25.6 to 9.1.
- The biggest single-benchmark swing is BigCodeBench Complete: 53.2% for DeepSeek-V2.5 (Sep 2024) and 43.1% for Llama 4 Scout.
Side by side
| DeepSeek-V2.5 (Sep 2024) | Llama 4 Scout | |
|---|---|---|
| Provider | DeepSeek | Meta |
| Noometry Index | 37.6 | 27.7 |
| Released | 2024-09-06 | 2025-04-05 |
| Weights | Open | Open |
| Context window | — | 128K |
| Max output | — | 4K |
| Input $ / M tokens | — | $0.10 |
| Output $ / M tokens | — | $0.30 |
| Results tracked | 22 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 31.7 (#281), Llama 4 Scout: 20.2 (#339)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 4 Scout |
|---|---|---|
| LMArena Coding | 1309 | 1286 |
| BigCodeBench Complete | 53.2% | 43.1% |
| SWE-bench Verified (bash only) | — | 9.1% |
| Aider Polyglot | 17.8% | — |
| SciCode | — | 17% |
| BigCodeBench Instruct | 48.6% | — |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Agentic & Tool Use Not comparable
DeepSeek-V2.5 (Sep 2024): —, Llama 4 Scout: 24.6 (#119)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 4 Scout |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 28.1% |
Reasoning DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 25.6 (#145), Llama 4 Scout: 9.1 (#345)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 4 Scout |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1266 |
| ARC-AGI-2 | — | 0% |
| Kagi LLM Benchmark | — | 36.9% |
| ARC-AGI-1 | — | 0.5% |
| CritPt | — | 0% |
| DTBench | — | 57.9% |
| LMCA | — | 12% |
| Epoch Capabilities Index | — | 129.64 |
| ForecastBench | — | 57.5 |
Math DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 35.9 (#177), Llama 4 Scout: 19.6 (#286)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 4 Scout |
|---|---|---|
| LMArena Math | 1288 | 1287 |
| OTIS Mock AIME 2024-2025 | — | 7.8% |
| Omni-MATH | — | 37.3% |
| MATH Level 5 | — | 62.3% |
| FrontierMath (Feb 2025 set) | — | 0% |
Knowledge DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 34.8 (#193), Llama 4 Scout: 31.9 (#217)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 4 Scout |
|---|---|---|
| LMArena Expert | 1266 | 1235 |
| GPQA Diamond | — | 51.8% |
| MMLU-Pro | — | 74.2% |
| Vectara Hallucination Rate | — | 7.7% |
| GPQA (HELM) | — | 50.7% |
Multimodal Not comparable
DeepSeek-V2.5 (Sep 2024): —, Llama 4 Scout: 32.2 (#102)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 4 Scout |
|---|---|---|
| LMArena Vision | — | 1118 |
| SpatialViz-Bench | — | 34.2% |
Multilingual DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 42.5 (#193), Llama 4 Scout: 41.0 (#212)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 4 Scout |
|---|---|---|
| LMArena Non-English | 1273 | 1252 |
| LMArena Chinese | 1318 | 1255 |
| LMArena French | 1289 | 1282 |
| LMArena German | 1258 | 1272 |
| LMArena Japanese | 1228 | 1206 |
| LMArena Korean | 1209 | 1207 |
| LMArena Russian | 1289 | 1263 |
| LMArena Spanish | 1248 | 1278 |
Instruction Following DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 67.5 (#194), Llama 4 Scout: 65.8 (#217)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 4 Scout |
|---|---|---|
| LMArena Instruction Following | 1280 | 1248 |
| IFEval | — | 81.8% |
Long Context DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#174), Llama 4 Scout: 27.5 (#294)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 4 Scout |
|---|---|---|
| LMArena Longer Query | 1301 | 1265 |
| Fiction.LiveBench | — | 36% |
Writing & Preference DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 49.8 (#187), Llama 4 Scout: 37.0 (#261)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 4 Scout |
|---|---|---|
| LMArena Text | 1294 | 1279 |
| LMArena Creative Writing | 1285 | 1249 |
| LMArena Multi-Turn | 1297 | 1280 |
| EQ-Bench Creative Writing | — | 783 |
| WildBench | — | 78% |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than Llama 4 Scout?
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.6 to 27.7 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or Llama 4 Scout better for coding?
DeepSeek-V2.5 (Sep 2024) scores higher on coding benchmarks: 31.7 versus 20.2 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Llama 4 Scout share?
18 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Llama 4 Scout has 43.