Back to GalleryBack
Model Comparison
Model Comparison
Compare performance, benchmarks, and characteristics
DeepSeek-V3
deepinfra
Context16K tokens
Input Price$0.49 / 1M tokens
Output Price$0.89 / 1M tokens
Loading comparison...
Model Comparison
| Metric |  |  |
|---|---|---|
Pricing | ||
| Input Price | $0.49 / 1M tokens | $2 / 1M tokens |
| Output Price | $0.89 / 1M tokens | $8 / 1M tokens |
Capabilities | ||
| Context Window | 16K tokens | 1047576 tokens |
| Capabilities | tools | tools |
| Input type | text | text, image |
Category Scores | ||
| Overall Average | 46.5 | 56.8 |
| Academia | 45.3 | 54.9 |
| Finance | 45.1 | 55.5 |
| Marketing | 49.4 | 54.4 |
| Maths | 57.0 | 67.5 |
| Programming | 35.6 | 41.9 |
| Science | 57.4 | 66.5 |
| Writing | 35.9 | 39.2 |
| Vision | N/A | 74.8 |
Benchmark Tests | ||
| AIME | 25.3 | 43.7 |
| AA Coding Index | 35.6 | 41.9 |
| AAII | 33.2 | 43.4 |
| AA Math Index | 57.0 | 67.5 |
| DROP | 91.6 | _ |
| GPQA | 57.4 | 66.5 |
| HLE | 3.6 | 4.6 |
| LiveCodeBench | 35.9 | 45.7 |
| MATH-500 | 88.7 | 91.3 |
| MMLU | 88.5 | 90.2 |
| MMLU-Pro | 75.6 | 80.6 |
| MMMU | _ | 74.8 |
| SciCode | 35.4 | 38.1 |
DeepSeek-V3 Advantages
- • Lower input token cost ($0.49 / 1M tokens vs $2 / 1M tokens)
gpt-4.1 Advantages
- • Higher benchmark scores (56.8 vs 46.5)
- • Larger context window (1047576 tokens vs 16K tokens)