Claude Opus 5 vs DeepSeek V4 Pro
Both are Flagship-tier models, from Anthropic and DeepSeek respectively. DeepSeek V4 Pro is the cheaper choice overall at a 4:1 input:output ratio.
Which is cheaper
- Input price: DeepSeek V4 Pro is cheaper ($1.32 vs $5.00).
- Output price: DeepSeek V4 Pro is cheaper ($3.96 vs $25.00).
- Blended, 4:1 input:output ratio: DeepSeek V4 Pro is 79.5% cheaper ($1.85 vs $9.00 per weighted 1M tokens).
Side by side
| Spec | Claude Opus 5 | DeepSeek V4 Pro |
|---|---|---|
| Provider | Anthropic | DeepSeek |
| Tier | Flagship | Flagship |
| Context window | 1M | 1M |
| Max output | 128K | 384K |
| Input $/1M | $5.00 | $1.32 |
| Output $/1M | $25.00 | $3.96 |
| Cache read $/1M | $0.50 | $0.044 |
| Cache write $/1M | $6.25 | — |
| Modalities | text, image | text |
Claude Opus 5. 5-minute cache write shown ($6.25 = 1.25x base input); 1-hour cache write is $10/MTok (2x base). Batch API is 50% off ($2.50/$12.50). Context window and max output cross-checked against platform.claude.com/docs/en/about-claude/models/overview.
DeepSeek V4 Pro. Peak cache-miss rate, which is this dataset's standard tier. DeepSeek's own page states off-peak rates are half the peak rates, with peak hours 01:00-04:00 and 06:00-10:00 UTC Monday to Friday: off-peak is $0.66 input / $1.98 output, cache-hit input $0.022. The off-peak half-price is a time-of-day discount and is excluded here for the same reason batch and cached-input discounts are.
Sources
Claude Opus 5 — verified August 29, 2026
DeepSeek V4 Pro — verified August 30, 2026
Full Claude Opus 5 pricing page → · Full DeepSeek V4 Pro pricing page →