llmrelay
// previous id vs current Flash

DeepSeek V4 Flash vs V4.1 Flash

V4 Flash is the old id. DeepSeek’s first-party docs now tell you to call deepseek-flash for V4.1 Flash. On this site the current id is deepseek-v4.1-flash. The old slug stays so existing configs keep working.

Previous generation

DeepSeek V4 Flash is still callable on llmrelay and still sold, but superseded. The comparison below is accurate; just know there is a newer generation if you are starting fresh.

DeepSeek V4 Flash
DeepSeek V4.1 Flash
Input / M (llmrelay)
$0.19
$0.19
Output / M (llmrelay)
$0.54
$0.54
Context window
1M tokens
1M tokens
Max output
384k tokens
384k tokens
Model id on llmrelay
deepseek-v4-flash
deepseek-v4.1-flash
DeepSeek status
Retired; remapped
GA 10 September 2026

Input / M (llmrelay): DeepSeek cache-miss list for V4.1 Flash is $0.15/$0.60 off-peak and $0.30/$1.20 peak. We bill one flat rate 24/7.

Model id on llmrelay: DeepSeek first-party uses deepseek-flash for V4.1. The expires-on-0910 preview id is dead.

DeepSeek status: api-docs.deepseek.com/updates, fetched 2026-09-11. First-party still accepts deepseek-v4-flash and routes it to V4.1 Flash.

Prices are llmrelay's, at 50% of official list. Specs are the vendor's own published figures (Anthropic model docs), not our benchmarks. We do not publish scores we cannot source.

Use deepseek-v4-flash when

  • +A config, eval set, or contract already names this exact id and you do not want to re-check output.
  • +You landed on the old URL from search and just need the same cheap Flash rate. It still bills $0.19/$0.54 here.

Use deepseek-v4.1-flash when

  • +You are starting new Flash work. This is the current id on llmrelay.
  • +You saw DeepSeek’s 10 September GA and want the current Flash, not the retired V4 Flash name.
  • +You tried the 09-08 expires-on-0910 preview. That name is gone. This is the replacement.

The honest answer

Same invoice, same upstream model. New work: deepseek-v4.1-flash. Keep deepseek-v4-flash only to pin an old config. One prepaid key covers both, plus Opus 5 and GPT-6 Astra.

What this costs you per month

At 50M input and 5M output tokens a month — a realistic heavy agent workload.

DeepSeek V4 Flash
$10.50
$12.20/mo
DeepSeek V4.1 Flash
$10.50
$12.20/mo

Struck-through column is the vendor's list price for the same traffic. Adjust the numbers on the calculator.