How much does the DeepSeek API cost?

It depends on the model. Across the 2 models tracked here, input ranges from $0.1400 to $0.4350 per million tokens and output from $0.2800 to $0.8700. On a typical chat turn of 1,500 input and 300 output tokens the range is $0.000294 to $0.000914 per request.

What is the cheapest DeepSeek model?

DeepSeek V4 Flash, at $0.1400 per million input tokens and $0.2800 per million output. Whether it is cheap enough depends on whether it can do your task — the cost of finding out is usually a few cents.

How current is this DeepSeek pricing?

Every model entry stores the URL it was read from and the date it was read, both shown in the table above. The full data set was last reviewed on August 5, 2026. Prices change without notice, so verify against DeepSeek's own documentation before committing spend.

2 models tracked

DeepSeek API pricing

DeepSeek is priced far below the other providers here, which makes it the reference point for whether a task genuinely needs a frontier model.

Models tracked: 2
Cheapest input: $0.1400
Highest input: $0.4350
Exact token counts: 0 / 2

Every DeepSeek model, cheapest first

DeepSeek model pricing
Model	Input	Cached	Output	Context	Count	Chat turn
DeepSeek V4 Flash	$0.1400	$0.002800	$0.2800	1M	Estimate	$0.000294
DeepSeek V4 Pro	$0.4350	$0.003625	$0.8700	1M	Estimate	$0.000914

Chat turn is 1,500 input and 300 output tokens. Prices read from DeepSeek’s pricing documentation; each row links to its own dated source.

What is distinctive about DeepSeek pricing

01
Cache hit pricing is dramatic — a small fraction of the cache miss rate — which rewards structuring prompts with a stable prefix more than any other provider on this list.
02
The gap between input and output pricing is narrower than at other providers, which changes the calculus for output-heavy work.
03
At these rates the cost of evaluating a task on DeepSeek is close to zero, so it is worth measuring rather than assuming a cheaper model cannot do the job.

Token counting on DeepSeek

DeepSeek ships a Hugging Face tokenizer, which can be used from a backend but is not bundled here. Counts on this site are therefore estimates.

Counts shown here for DeepSeek models are estimates and are labelled as such everywhere they appear. The methodology page documents every scaling factor used.

Compare with other providers

Or shortlist across all of them at once in the model finder, filtering by budget, context window and whether token counts can be exact.

Frequently asked questions

How much does the DeepSeek API cost?: It depends on the model. Across the 2 models tracked here, input ranges from $0.1400 to $0.4350 per million tokens and output from $0.2800 to $0.8700. On a typical chat turn of 1,500 input and 300 output tokens the range is $0.000294 to $0.000914 per request.
What is the cheapest DeepSeek model?: DeepSeek V4 Flash, at $0.1400 per million input tokens and $0.2800 per million output. Whether it is cheap enough depends on whether it can do your task — the cost of finding out is usually a few cents.
Can I count DeepSeek tokens exactly?: Not in a browser. DeepSeek ships a Hugging Face tokenizer, which can be used from a backend but is not bundled here. Counts on this site are therefore estimates. Counts shown on this site for DeepSeek models are labelled as estimates everywhere they appear.
How current is this DeepSeek pricing?: Every model entry stores the URL it was read from and the date it was read, both shown in the table above. The full data set was last reviewed on August 5, 2026. Prices change without notice, so verify against DeepSeek's own documentation before committing spend.