Claude Sonnet 5.5 vs DeepSeek V4 Pro 0813: cost per task
DeepSeek V4 Pro 0813 is cheaper on 12 of 14 tasks and about 2.3x cheaper on average (geometric mean).
| Task | Claude Sonnet 5.5 | DeepSeek V4 Pro 0813 | Difference |
|---|---|---|---|
| Answer a chat question | 1.3¢ | 1.3¢ | 1.0x |
| Reply to a customer email | 0.40¢ | 0.41¢ | 1.0x |
| Write a 1,500-word blog post | 3.7¢ | 0.61¢ | 6.0x |
| Write 10 product descriptions | 4.4¢ | 0.56¢ | 7.9x |
| Translate 1,000 words | 2.6¢ | 2.4¢ | 1.1x |
| Proofread one page (500 words) | 1.1¢ | 0.57¢ | 2.0x |
| Summarize a 20-page PDF | 4.5¢ | 1.3¢ | 3.6x |
| Extract data from an invoice | 1.0¢ | 0.60¢ | 1.7x |
| Analyze a 200-row spreadsheet | 17¢ | 3.5¢ | 4.9x |
| Answer from 5 retrieved documents (RAG) | 2.7¢ | 0.99¢ | 2.8x |
| Classify 100 support tickets | 0.89¢ | 1.1¢ | 1.2x |
| Review 500 lines of code | 8.0¢ | 3.5¢ | 2.3x |
| Write a SQL query | 1.1¢ | 0.43¢ | 2.5x |
| Run a 50-step coding agent | 84¢ | 30¢ | 2.8x |
Specs and list prices
| Claude Sonnet 5.5 | DeepSeek V4 Pro 0813 | |
|---|---|---|
| Input per 1M | $2.00 | $0.66 |
| Output per 1M | $10.00 | $1.98 |
| Cached input per 1M | $0.10 | $0.022 |
| Context | 1M | 1M |
| Providers | 8 | 23 |
| Reasoning | always | default |
| LMArena Text | 1475 | - |
| Released | Sep 28, 2026 | Aug 12, 2026 |
FAQ
Is Claude Sonnet 5.5 or DeepSeek V4 Pro 0813 cheaper?
DeepSeek V4 Pro 0813 is cheaper on 12 of 14 tasks and about 2.3x cheaper on average (geometric mean).
How much does Claude Sonnet 5.5 cost compared to DeepSeek V4 Pro 0813 per token?
Claude Sonnet 5.5: $2.00 input and $10.00 output per 1M tokens. DeepSeek V4 Pro 0813: $0.66 input and $1.98 output. Token prices alone mislead - models use different numbers of tokens for the same task, which is what the task table shows.