LLMOpen weights

DeepSeek-V4-Flash

DeepSeek

Updated May 23, 2026 · Benchmarks via Artificial Analysis, specs & pricing via OpenRouter · Methodology · Report an error

Specifications

Type: LLM
Access: Open weights
Released: April 24, 2026
License: MIT
Parameters: 284B (13B active)
Context window: 1,048,576 tokens
Output speed: 109 tok/s
Latency (TTFT): 0.76s
Input: Text
Output: Text
API pricing: $0.10 in · $0.20 out / 1M tokens

Capabilities

Function callingStructured output

Benchmarks

Reasoning

89.4

General

Humanity’s Last Exam

32.1

Our take

Smaller, cheaper sibling of V4-Pro with the same native 1M context — the cost-efficient tier of the V4 series.

Links

Model card Announcement

Compare DeepSeek-V4-Flash with

vs Sonar Reasoning Pro vs R1 1776 vs Qwen3.7 Max vs Gemini 3.5 Flash

See all DeepSeek-V4-Flash alternatives →