DeepSeek · Workhorse

DeepSeek-V4-Pro-0813 API pricing and specs

DeepSeek's larger text-only model, kept in service after a planned retirement was reversed.

Input

⁨$1.32⁩

Cached input

⁨$0.044⁩

Output

⁨$3.96⁩

US dollars per million tokens, standard tier, short-prompt rate

Thinking is on by default with low, high and max effort. It has no image input.

DeepSeek charges half price off-peak. Peak hours are 01:00 to 04:00 and 06:00 to 10:00 UTC on weekdays; every other hour, weekends included, is off-peak. The table shows peak prices.

Spec sheet

API model string: deepseek-v4-pro
Provider: DeepSeek
Context: 1M tokens
Max output: 384K tokens
Released: August 13, 2026
Batch API: No batch discount listed
Documented agent features: Tool callingStructured outputAdjustable reasoning

What a task costs

Four reference workloads priced at this model's list rates, with caching applied after the first call. Your numbers will differ; the calculator lets you change every assumption.

Workload	Per task	Per 1,000 tasks
Support bot 4 calls, short tool lookups	⁨$0.029⁩	⁨$29⁩
Coding agent 30 calls, files and test output	⁨$0.961⁩	⁨$961⁩
Browser agent 25 calls, screenshots and page text	⁨$0.822⁩	⁨$822⁩
Research agent 15 calls, long search results	⁨$0.771⁩	⁨$771⁩