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 | 2026-07-31 |
| Parameters (B) | 284 |
| Architecture | Mixture-of-Experts Transformer |
| Context Window | 1000000 |
| Input Modalities | text |
| Open Source | No |
| License | |
| Score | 0 |
| VRAM (GB) | 192 |
| Compute | 2x NVIDIA H200 141GB or single AMD MI300X 192GB |
| Benchmark: AIME-2026 | 95.8 |
| Benchmark: Agents-Last-Exam | 25.2 |
| Benchmark: AutomationBench | 37.7 |
| Benchmark: BrowseComp | 73.2 |
| Benchmark: CodeForces | 3289 |
| Benchmark: CyberGym | 38.7 |
| Benchmark: DeepSWE | 54.4 |
| Benchmark: GDPval-AA | 1395 |
| Benchmark: GPQA-Diamond | 89.9 |
| Benchmark: HLE | 37.8 |
| Benchmark: IFEval | 79.2 |
| Benchmark: IMO-AnswerBench | 88.4 |
| Benchmark: LiveCodeBench | 91.6 |
| Benchmark: MMLU-Pro | 86.4 |
| Benchmark: NL2Repo-Bench | 39.4 |
| Benchmark: SWE-Bench-Multilingual | 73.3 |
| Benchmark: SWE-Bench-Pro | 52.6 |
| Benchmark: SWE-Bench-Verified | 79 |
| Benchmark: Terminal-Bench-2.0 | 56.9 |
| Benchmark: Terminal-Bench-2.1 | 82.7 |
| Benchmark: Terminal-Bench-3.0 | 7.6 |
| Benchmark: Toolathlon-Verified | 49.7 |
| Pricing (Input) | 0.00 |
| Pricing (Output) | 0.35 |