← All open-weight models

DeepSeek · General language

DeepSeek-V4-Flash

The default DeepSeek pick: near-Pro reasoning at a third of the price with the same 1M context, ideal for high-volume agentic coding and long-document work — schedule batchy jobs outside 01:00-04:00/06:00-10:00 UTC weekdays and you pay $0.22/$0.66 instead of $0.44/$1.32.

Open weights · Permissive (Apache, MIT, OpenMDW) · Data checked September 6, 2026 · Not a hands-on evaluation

Model ID deepseek-v4-flash

Parameters
304B total · 13B active304B total MoE (~13B active; V4 preview quoted 284B total / 13B active, architecture unchanged in the 0731 re-post-train)
Architecture
Mixture of experts
Context window
1,000,000 tokensMaximum input
Max output
384,000 tokensPer response
Input price
$0.44 per 1M tokens
Output price
$1.32 per 1M tokens
Cached input
$0.014 per 1M tokens
Blended price
$0.66 per 1M tokens3:1 input to output
Knowledge cutoff
Not recorded
Released
2026-07-31
Status
gaFlagship in its lineup
Size band
120B and above

License and openness

DeepSeek-V4-Flash is released under the MIT. This is a permissive license. It generally allows use, modification and redistribution, including commercial use, subject to attribution and notice requirements. License terms can change between versions; confirm the text published with the weights you download.

Capabilities

Input modalitiestext
Output modalitiestext

What it is for

At a three-to-one input-to-output ratio, DeepSeek-V4-Flash costs $0.66 per million tokens blended. A workload of one million input and 330,000 output tokens per day would run about $26.27 per month at list price, before caching or batch discounts.

Cost at list price

Input$0.44 per 1M tokens
Output$1.32 per 1M tokens
Cached input$0.014 per 1M tokens
Blended, 3:1$0.66 per 1M tokens
Estimate a workload across all models ↗

Sources

Sources checked September 6, 2026. Specifications and prices change without notice; confirm against the provider before you commit.

Other open-weight models from DeepSeek

Browse all open-weight models ↗