š Introducing DeepSeek-V3.2-Exp ā our latest experimental model! ⨠Built on V3.1-Terminus, it debuts DeepSeek Sparse Attention(DSA) for faster, more efficient training & inference on long context. š Now live on App, Web, and API. š° API prices cut by 50%+! 1/n
- Views
- 1,424,571
- Likes
- 6,919
- Replies
- 321
- Reposts
- 871
- Quotes
- 354
- Bookmarks
- 1,055
š Introducing DeepSeek-V3.2-Exp ā our latest experimental model!
DeepSeek announces an experimental model built on V3.1-Terminus. The post highlights a new sparse-attention feature, availability across App, Web, and API, and a price reduction of more than 50% for API use.
- Views
- 1.4M
- Likes
- 6.9K
- Replies
- 321
- Reposts
- 871
- Quotes
- 354
- Bookmarks
- 1.1K
- Topics
Analysis
DeepSeek announces an experimental model built on V3.1-Terminus. The post highlights a new sparse-attention feature, availability across App, Web, and API, and a price reduction of more than 50% for API use.
Sign in to see the full analysis.
Sign inš Introducing DeepSeek-V3.2-Exp ā our latest experimental model! ⨠Built on V3.1-Terminus, it debuts DeepSeek Sparse Attention(DSA) for faster, more efficient training & inference on long context. š Now live on App, Web, and API. š° API prices cut by 50%+! 1/n
š Introducing DeepSeek-V3.2-Exp ā our latest experimental model!
- 01The Announcement
Introduce the new offer with energy and frame it as a fresh or experimental release.
- 02The Main Benefit
Explain what the offer adds and connect it to a clear improvement in performance or efficiency.
- 03Availability and Value
Tell people where they can use it and highlight a concrete price or access advantage.
- 04The Continuation Marker
Signal that the announcement is part of a longer thread or sequence.