DeepSeek V4 Pro 0813 API Prices Double: Input Now $1.32, Output $3.96 per 1M Tokens

What Changed

DeepSeek quietly doubled the API pricing for V4 Pro 0813. Input tokens moved from $0.66 to $1.32 per 1M tokens, and output tokens jumped from $1.98 to $3.96 per 1M tokens — a clean +100% increase on both sides.

See the full pricing timeline on the DeepSeek V4 Pro 0813 — live specs & price history page.

Does This Matter for Your Workload?

It depends on your token ratio. Output-heavy workloads — long-form generation, summarization, code completion — take the harder hit since output was already priced at 3× input, and that gap is now amplified in absolute terms. At $3.96 per 1M output tokens, a pipeline burning 10M output tokens monthly goes from roughly $19.80 to $39.60.

For read-heavy or classification tasks with short outputs, the impact is more modest, but the input cost still doubled, so no workload is insulated.

What to Do Next

• Audit your token split. If you're output-heavy, reconsider prompt design to trim generation length.
• Compare alternatives. Several models in our catalogue sit in the $1–$4 output range with competitive benchmarks. Filter by price on the models page to find candidates worth testing.
• Check for caching. If your provider or integration layer supports prompt caching, this is a good moment to enable it — input savings compound quickly at $1.32/1M.

Bottom Line

A 100% price increase is significant, not routine. DeepSeek V4 Pro 0813 may still be cost-competitive depending on quality-per-dollar in your specific use case, but the calculus just shifted enough to warrant a fresh benchmark-vs-cost review before your next billing cycle.

Ezra, Scout AI Team