Ask
Deprecation

DeepSeek reverses the V4-Pro retirement it announced a day earlier

DeepSeek pricing

DeepSeek withdrew its 2026-09-14 plan to route deepseek-v4-pro traffic to V4.1-Flash. V4-Pro keeps its own API service and unchanged rates from $0.66/1M.

Before

Pricing-page footnote (2026-09-10): from 12:00 Beijing Time on 2026-09-14, and until V4.1 Pro is released, all requests to deepseek-v4-pro would be routed to V4.1 Flash and billed at the V4.1 Flash price.

After

Pricing-page footnote (2026-09-11): API services for DeepSeek V4 Pro continue after September 14, 2026, with the billing method remaining unchanged. V4-Pro keeps its own row at $0.022 cache-hit / $0.66 cache-miss / $1.98 output per 1M off-peak, double at peak.

Nothing on the rate card moved, which is the point. One day after announcing that DeepSeek-V4-Pro would be retired in an orderly manner — with all deepseek-v4-pro requests routed to V4.1-Flash and billed at the cheaper Flash price from 12:00 Beijing Time on 2026-09-14 — DeepSeek replaced that footnote with its reversal: “In response to user demand, we have decided to continue providing API services for DeepSeek V4 Pro after September 14, 2026, with the billing method remaining unchanged. We will provide further notice should there be any changes.”

For teams on V4-Pro this removes a three-day forced-migration deadline and keeps the higher-capacity model available at its existing rates, but it also underlines how DeepSeek handles lifecycle: retirement dates are published as pricing-page footnotes and can be added or withdrawn overnight, with no versioned deprecation policy and no contractual notice period. Anyone who started a migration off V4-Pro on the strength of the 2026-09-10 footnote did so against a commitment that lasted a day.

From DeepSeek's pricing timeline
DeepSeek-V4.1-Flash Launches — Flash Rates Cut Across the Board

DeepSeek released DeepSeek-V4.1-Flash, called as the model name deepseek-flash, and cut every Flash line item on the peak/off-peak card: cache-hit input from $0.007 to $0.003 off-peak ($0.014 to $0.006 peak), cache-miss input from $0.22 to $0.15 off-peak ($0.44 to $0.3 peak), and output from $0.66 to $0.6 off-peak ($1.32 to $1.2 peak). The separate DeepSeek-V4-Flash-0731 and DeepSeek-V4-Flash-Vision-Exp rows were retired into it — vision is now a feature of Flash, and the legacy names still route to V4.1-Flash at the Flash price. V4-Pro's six rates were unchanged, and its announced 2026-09-14 retirement was reversed on 2026-09-11.

About DeepSeek
api-docs.deepseek.com ↗

DeepSeek offers a free web chat product and a pay-per-token API, billed since 2026-08-16 on a peak/off-peak rate card: DeepSeek-V4.1-Flash (general-purpose, from $0.003/1M cache-hit input off-peak / $0.006 peak, $0.15 cache-miss in off-peak / $0.3 peak, $0.6 out off-peak / $1.2 peak) and DeepSeek-V4-Pro (higher-capacity, from $0.022/1M cache-hit input off-peak / $0.044 peak, $0.66 cache-miss in off-peak / $1.32 peak, $1.98 out off-peak / $3.96 peak) — both with a 1M-token context window and dramatically cheaper than equivalent OpenAI and Anthropic models.

Pricing model freemiumpure usage
Billing units tokensapi calls
Sales motion self serveplg
Free tier
Yes
Commits
None
Transparency
public

DeepSeek pricing history

  1. Sep 2026
    DeepSeek-V4.1-Flash Launches — Flash Rates Cut Across the Board
  2. Aug 2026
    DeepSeek Announces Peak/Off-Peak Pricing — Effective Aug 16, 2026
  3. Mar 2025
    DeepSeek-V3-0324 Update — Improved Coding
  4. Jan 2025
    Nvidia Stock Drops 17% Following R1 Release
  5. Jan 2025
    DeepSeek-R1 Released — Reasoning Model, MIT Open-Source
Full DeepSeek timeline

More DeepSeek activity

All pricing activity