Claude 3.7 Sonnet vs DeepSeek V3

Per-platform pricing

Each row retains its own currency, unit, region, and source status. Rates with different scopes are not ranked against each other.

Frequently asked questions

Which is cheaper, Claude 3.7 Sonnet or DeepSeek V3?

DeepSeek V3's stored input rate is 91% lower than Claude 3.7 Sonnet's ($0.27 vs $3.00 per 1M Tokens; matched scope: 1M Tokens · global · llm). DeepSeek V3's stored output rate is 93% lower than Claude 3.7 Sonnet's ($1.10 vs $15.00 per 1M Tokens; matched scope: 1M Tokens · global · llm).

Where does this comparison data come from?

This page uses stored price records. A provider or platform source is linked where the record contains one; records without a source link are explicitly marked. Record timestamps show database updates, not independent verification.

What does this comparison include?

It compares a price field only when both records have the same currency, billing unit, region, and modality. It does not normalize taxes, volume discounts, availability, latency, quality, or provider-specific terms.