DeepSeek

Flash model version updated from DeepSeek-V4.1-Flash to DeepSeek-V4.1-Flash-1017; token prices unchanged. New standalone Search and Fetch APIs priced at $0.75 per 1,000 successful requests each; currently free via Harness/Claude Code/Codex, per-request billing starts Nov 1, 2026.

Weekly roll-up · includes Oct 17 daily

New Search and Fetch APIs priced at $0.75 per 1,000 successful requests each; server-side tool calls free until per-request billing starts Nov 1, 2026.

Peak/off-peak billing window redefined: Chinese public holidays are now explicitly off-peak (50% discount), previously they fell under normal Mon-Fri peak-hour pricing.

Weekly roll-up · includes Sep 20 daily

Off-peak/peak hours definition changed: peak windows (01:00-04:00 and 06:00-10:00 UTC weekdays) now explicitly exclude Chinese public holidays, so holidays get the 50% off-peak discount all day like weekends.

DeepSeek merged v4-flash and vision-exp into one deepseek-flash (V4.1) model at lower cache-hit/miss and output prices; new footnote confirms V4 Pro pricing stays unchanged past Sept 14, 2026.

Weekly roll-up · includes Sep 12 daily, Sep 11 daily

DeepSeek reversed the planned Sept 14 V4 Pro retirement: footnote now says Pro API continues with billing unchanged, replacing the prior notice that Pro requests would reroute to V4.1 Flash pricing.

New notice: DeepSeek will retire the V4 Pro model starting Sept 14, 2026 12:00 Beijing Time, rerouting all deepseek-v4-pro requests to V4.1 Flash and billing them at the (cheaper) Flash price until a V4.1 Pro model ships. Flash tier token prices cut across the board (cache-hit input -57%, cache-miss input -32%, output -9%) while the separate flash-vision-exp model was retired and merged into the base Flash model, which now natively supports Vision. Pro tier pricing unchanged.

Peak-hours footnote now specifies peak pricing applies Monday through Friday only, so weekends are always billed at off-peak rates; previously the schedule applied every day. New deepseek-v4-flash-vision-exp model added to the pricing table, priced identically to deepseek-v4-flash ($0.22 off-peak / $0.44 peak per 1M cache-miss input tokens, $0.66 off-peak / $1.32 peak per 1M output tokens, 2500 concurrency); a new footnote states images are converted to tokens by dimension and billed as input tokens, and FIM Completion is not supported.

DeepSeek API token pricing restructured from flat rates to peak/off-peak billing (peak 01:00-04:00 & 06:00-10:00 UTC) with a significant increase across both models: v4-flash cache-hit $0.0028→$0.007/$0.014, cache-miss $0.14→$0.22/$0.44, output $0.28→$0.66/$1.32; v4-pro cache-hit 0.003625→$0.022/$0.044, cache-miss $0.435→$0.66/$1.32, output $0.87→$1.98/$3.96 (off-peak/peak per 1M tokens)

DeepSeek replaced its announced peak/off-peak 2x pricing policy notice with a general warning of a significant overall API price increase coming soon. Current listed per-token prices are unchanged.

Announced upcoming peak/off-peak pricing policy: prices will be 2x the regular rate during peak hours (9:00-12:00 and 14:00-18:00 Beijing time, daily), applicable to all billing items; effective date not yet set