ModelPriceWatch.com
Last scan 2026-09-22 Models tracked 267 Providers 35 Cheapest paid Granite 4.0 H Micro $0.017/Mtok in Every price links to its source

DeepSeek

Commercial provider · 2 models tracked · Founded 2023

Chinese AI lab offering extremely cheap APIs with aggressive prompt caching. From 2026-08-16 it bills peak/off-peak by the hour, with off-peak at half the peak rate — see each model row for both.

DeepSeek pricing at a glanceSeptember 2026 · $ per 1M tokens

DeepSeek API pricing (September 2026): 2 current models range from $0.300 to $1.32 per 1M input tokens and $1.20 to $3.96 per 1M output tokens. The cheapest paid model is DeepSeek V4.1 Flash at $0.300/1M input; the priciest is DeepSeek V4 Pro at $1.32/1M input / $3.96 output. Every price links to DeepSeek's official pricing page and refreshes twice daily.

Models2
Input range$0.300–$1.32
Output range$1.20–$3.96
Cached-input tiers2

Today's DeepSeek prices

2 models · cheapest blended first

Sorted by blended cost (cheapest first). Prices per 1M tokens, September 2026 — every price links to DeepSeek's official pricing page.

Current DeepSeek model prices per 1M tokens, sorted by blended cost
Model Blended* Input Output Cached in Relative cost Context Status
DeepSeek V4.1 Flash
$0.525
$0.300 $1.20 $0.006
1M Current
DeepSeek V4 Pro
$1.98
$1.32 $3.96 $0.044
1M Current
* Blended = (3×input + 1×output) ÷ 4 $/1M tokens · cheap · mid · expensive MODELPRICEWATCH.COM · 2026-09-22

Quick stats

Models tracked
2
Type
commercial
Founded
2023
Cheapest model
DeepSeek V4.1 Flash
Cheapest blended
$0.525/M

Try DeepSeek

Sign up and start building with DeepSeek models.

Get started