Compare LLM Models
Select 2-5 models to compare side by side.
Select models to compare
0 / 5 selected
Top downloads
Select 2-5 models
| Attribute |
deepseek
|
|---|---|
| Company | deepseek |
| Release Date | 2025-10-09 |
| Parameters (B) | 552 |
| Architecture | Mixture-of-Experts (MoE) Transformer |
| Context Window | 1000000 |
| Input Modalities | text, image |
| Open Source | No |
| License | |
| Score | 39 |
| VRAM (GB) | — |
| Compute | API only |
| Benchmark: AGIEval | 83.4 |
| Benchmark: AIME-2025 | 87.5 |
| Benchmark: Agents-Last-Exam | 31.8 |
| Benchmark: AutomationBench | 54.8 |
| Benchmark: BBH | 86.1 |
| Benchmark: C-Eval | 92.1 |
| Benchmark: CodeForces | 3471 |
| Benchmark: CyberGym | 88.1 |
| Benchmark: DROP | 87.9 |
| Benchmark: DeepSWE | 74.2 |
| Benchmark: GPQA-Diamond | 90.9 |
| Benchmark: GSM8K | 93 |
| Benchmark: HLE | 39.1 |
| Benchmark: HellaSwag | 87.2 |
| Benchmark: HumanEval | 79.4 |
| Benchmark: IFEval | 89.5 |
| Benchmark: LiveBench | 81.1 |
| Benchmark: MATH | 61.1 |
| Benchmark: MMLU | 91 |
| Benchmark: MMLU-Pro | 81.2 |
| Benchmark: MMMU-Pro | 56.5 |
| Benchmark: NL2Repo-Bench | 65.4 |
| Benchmark: SimpleBench | 66.7 |
| Benchmark: Terminal-Bench-2.1 | 90.6 |
| Benchmark: Terminal-Bench-3.0 | 30 |
| Pricing (Input) | 0.11 |
| Pricing (Output) | 0.34 |