Model board / DeepSeek-V4-Flash

DeepSeek-V4-Flash

Name: DeepSeek-V4-Flash
Brand: DeepSeek
Price: 1.32 USD

latest DeepSeek

Verified 2026-09-03 against the pages listed below. Prices are USD per 1M tokens, list rate — no committed-use, enterprise or volume discount.

Input

$0.44

Cached input

$0.014

Cache write

—

Output

$1.32

Context window

1,000,000

Max output

384,000

Weights

open

Caching

automatic

list rate is the peak rate: DeepSeek's own pages call peak standard and off-peak a 50% discount, so this board has shown $0.44/$0.014/$1.32 since 2026-08-27, where it previously showed the off-peak $0.22/$0.007/$0.66. Nothing about DeepSeek's rates moved this pass — only which of the two this board puts in the price column. Off-peak is every hour outside 01:00–04:00 and 06:00–10:00 UTC, Monday to Friday

cache hit $0.007 vs miss $0.22 off-peak (~3%); reasoning effort is now documented in three named levels — low, high, max — rather than a flat 'controllable', and the API added native OpenAI Responses API support with one-click Codex integration

Licence: MIT (reported)

Watch out

Priced here at the peak/standard rate. Off-peak halves it, so a job you can schedule outside 01:00–04:00 and 06:00–10:00 UTC bills at half this row — and one you cannot schedule bills at this row, not the number quoted in most comparisons

What we would use it for

Bulk extraction, self-hosting, fine-tuning — still a 1M window at a fraction of frontier price

What this has cost, week by week

Output price at each refresh since 2026-07-30. A vendor change is the provider changing its price. Our correction means this board changed what it prints and the provider's charge did not move — the two are never merged, because publishing the second as the first would be a price rise that never happened.

07-30 $0.28→08-06 $0.28→08-14 $0.28→08-20 $0.66→08-27 $1.32→09-03 $1.32

vendor change2026-08-20 · in, cached_in, out — The peak/off-peak change announced 2026-08-13 went live on schedule at 2026-08-16 16:00 UTC. At this pass the board printed the off-peak rate — see the 2026-08-27 correction.

our correction2026-08-27 · in, cached_in, out — Re-based from the off-peak rate to the peak rate, which DeepSeek's own pages call the standard rate. DeepSeek charged the same before and after; the board's column header promises list rates and was printing a time-of-day discount.

Where these numbers came from

No number on this page was written that was not read from one of those pages on 2026-09-03. If one is wrong, send us the source URL and it is corrected in the next weekly refresh, with a changelog line saying what was wrong — contact@naderu.com.

Jobs this is one of our picks for

Extraction — Classify or pull fields out of a corpus (extraction)
Summarisation — Long documents in, short faithful text out (summarisation)

Discussed in Naderu Weekly

The clock is now part of the bill — 2026-08-14

The rest of the board

This row sits alongside 53 others on the model board, refreshed every week. The same data as JSON, and the terms for using it, are at /board-data.