All models

DeepSeek-V4-Flash

byDeepSeekDeepSeek· 24 Apr 2026
General purpose

DeepSeek has finally released its successor to the V3 series, which it kept updating for months. It comes in two sizes: Pro, which is a 1.6T-A49B MoE, and Flash, a 284B-13B model. Based on others' experience, the latter model seems to be the real star of the show, as its performance is relatively strong, while Pro seems to underdeliver relative to its size. The tech report goes into great detail, including the architectural changes used to achieve better and cheaper long-context performance.

Specs
Params284B (13B active)
LicenseMIT
Similarity · VAIL
VAIL
VAIL Fingerprint
0467:0567:0844:0acc:0ecd:13cb:20d0:51e0

Explore other models with behavioral similarity to DeepSeek-V4-Flash.

Capability · Artificial Analysis
AA Index
42.1
Months Behind Frontier
4.4 mo
gpt-5-2
11 Dec 2025
Adoption · Hugging Face
RAM score
Relative Adoption Metric not applicable.
Hugging Face Downloads
1.8M
last 30d
10.4M
all time
HF Likes
2.1K

Relative Adoption Metric not applicable.

Inference · OpenRouter
Tokens/Day
756.7B
7d avg · listed 7/7
Peak Tokens/Day
1.2T
29 Jul 2026
Peak Rank
#1
15 May 2026

OpenRouter publishes daily token totals for its ~50 most-served models. Days missing from the chart mean DeepSeek-V4-Flash fell below that cutoff — not zero usage. OpenRouter logs usage separately per dated model version and per variant (like ":free"); the hub combines provider variants while keeping separately cataloged releases distinct.

Related Models