Based on weekly OpenRouter usage tracked by AI Tier List, DeepSeek: DeepSeek V4 Pro ranks #6 overall — ahead of Anthropic: Claude Sonnet 5 at #13. DeepSeek: DeepSeek V4 Pro is cheaper on input tokens at $0.43/1M vs $2/1M for Anthropic: Claude Sonnet 5. DeepSeek: DeepSeek V4 Pro offers the larger context window at 1M tokens.
| Anthropic: Claude Sonnet 5 | DeepSeek: DeepSeek V4 Pro | |
|---|---|---|
| Author | anthropic | deepseek |
| Input price (per 1M tokens) | $2 | $0.43 |
| Output price (per 1M tokens) | $10 | $0.87 |
| Context window | 1M | 1M |
| Input modalities | text, image, file | text |
| Output modalities | text | text |
| AA Intelligence | 53.4 | 44.3 |
| AA Coding | 71.5 | 59.4 |
| AA Agentic | 46.7 | 36.4 |
| Released | Jun 2026 | Apr 2026 |
| Task | Anthropic: Claude Sonnet 5 | DeepSeek: DeepSeek V4 Pro |
|---|---|---|
| Overall | #13 | #6 |
| Memory Extraction | — | #7 |
| Multi-step Planning | — | #5 |
| Tool Dispatch | — | #4 |
| Web Search | — | #4 |
| Agent Workflows | — |
Usage ranks are weekly data aggregated by OpenRouter. Prices are USD per 1M tokens.
| #7 |
| Classification & Tagging | — | #10 |
| Debugging | — | #5 |
| DevOps Config | — | #6 |
| File Read/Write | — | #7 |
| Frontend UI | — | #9 |
| Code Implementation | — | #5 |
| Repo Analysis | — | #7 |
| Code Review & Security | — | #6 |
| Shell Execution | — | #6 |
