Skip to content

DeepSeek-V3

DeepSeek's language model built on a mixture-of-experts architecture.

Find in search
3
mentions in the archive
0
central to an episode
1
episode with a mention
1
fragment with a timecode
Apr 2025
first mention
Apr 2025
peak · 1 in a month
16
links to other entries
Parent companyDeepSeekCompany · Hangzhou, China64episodes0products2modelsCompany page →

Place in the ecosystem

DeepSeek-V3Model
also in the lineupDeepSeek-R1

Model card

DeveloperDeepSeek
Used in—
First mention30 Apr 2025
Episodes where it is central0
Spellings in the indexdeepseek-v3deepseek v3v3 deepseek
NOTEDeepSeek's language model built on a mixture-of-experts architecture.

Author's trends

In the episodes of the author's 20 thoughts, DeepSeek-V3 is never central or discussed in depth, and the author's quotes don't mention the name.
MAIN EPISODEspecial · 21:26
special·30 Apr 2025context and comparison✓ article
“Open Source” Does Not Mean “Do Anything You Want”: How to Read AI Model Licenses

Mentions over time

202420252026

Key fragments