Modelpeers
DeepSeek-V4.1-Flash character

ModelDeepSeek · Joined Sep 10, 2026

DeepSeek-V4.1-Flash

DeepSeek's smallest model in its new architecture family, a 552B-parameter MoE with native visual understanding.

Available for DeepSeek API

Compare
Official pageReport a correction

The basics

Checked Oct 31
Status
ActiveLive on the DeepSeek API with native multimodal support.
Model ID

Updates

Security

DHS orders US federal civilian agencies to remove DeepSeek products

A Department of Homeland Security order published in the Federal Register on October 20, 2026 tells federal civilian agencies to stop buying DeepSeek products, including third-party products built with DeepSeek technology, and remove them from their systems. Agencies can request exceptions within 60 days.

Action needed:
If you sell to US federal civilian agencies, check whether your product uses DeepSeek models.
Full changelog, by month

Strengths and weaknesses

Strengths

  • Native visual understanding

Weaknesses

  • !

    Running shell commands reliably

Role

Open-weight model at DeepSeek

  1. DeepSeek-V4.1-Flash character

    DeepSeek-V4.1-FlashCurrent

    Sep 2026 to now · 2 months
  2. DeepSeek-V4-FlashApr 2026 to Sep 2026 · 5 months

Works with

From recorded changes
DeepSeek API

API · 3 shared changes

DeepSeek Harness

Agent · 1 shared change

Gemini Enterprise Agent Platform

API · 1 shared change

Similar models

DeepSeek-V4-Pro character
DeepSeek-V4-Pro

$1.32 in, $3.96 out

DeepSeek-V4-Flash

$0.15 in, $0.60 out

DeepSeek-V4-Flash-Vision-Exp

$0.15 in, $0.60 out

Similar models for now. Real viewing data comes after launch.

Where it works

PlacePrice per 1M tokens, in / outUptime
OpenInference$0.01 / $0.1499.93%
Baidu$0.04 / $0.1898.89%
Decart$0.09 / $0.18100%
Morph$0.02 / $0.1999.86%
Sail Research$0.09 / $0.25100%
DeepInfra$0.11 / $0.3499.85%
Wafer$0.04 / $0.50100%
Relace$0.01 / $0.5397.7%

Prices and uptime are OpenRouter's figures for each provider's cheapest endpoint. Uptime covers the last 24 hours.

Benchmarks

Higher is better. Green is DeepSeek-V4.1-Flash. Dark marks are similar models.

Independent

What people say

Reviews

No reviews yet. Be the first to review DeepSeek-V4.1-Flash.

Used DeepSeek-V4.1-Flash? Tell other builders what worked and what did not.

Incidents

Writing habits

Creative writing · EQ-Bench, among 51 models tracked here

Slop score

22.4

Overused AI words and phrases. Less slop than 28% of tracked models.

Repetition

5.2

Repeated words and ideas. Less repetition than 14% of tracked models.

Story length

6,121

Average characters per story. Longer than 42% of tracked models.

Research sources

Researched Oct 31, 2026. Each section's own sources open with its Show sources button.