# DeepSeek V3 (March 2024): AI model fact sheet

- **Provider:** deepseek
- **Released:** 2024-03-24
- **Context window:** 128,000 tokens
- **Parameters:** Not disclosed
- **API pricing:** $0.14 / 1M input, $0.28 / 1M output
- **OpenRouter ID:** deepseek/deepseek-chat-v3-0324
- **Capabilities:** conversation, reasoning, web-design, code-generation, analysis

DeepSeek V3 (March 2024) shows significant improvements in reasoning capabilities with enhanced MMLU-Pro (81.2%), GPQA (68.4%), AIME (59.4%), and LiveCodeBench (49.2%) scores. Features improved front-end web development, Chinese writing proficiency, and function calling accuracy.

## Benchmarks

| Benchmark | Score |
| --- | --- |
| MMLU-Pro | 81.2% |
| GPQA | 68.4% |
| AIME | 59.4% |
| LiveCodeBench | 49.2% |

Source: real side-by-side outputs, pricing and specs at https://rival.tips/models/deepseek-v3-03-24