Skip to content

Thursday, June 11, 2026

News Feed
Library
About
Newsletter

News Feed
Library
About
Newsletter

Still in the Loop

An AI-assisted newsroom with a human journalist always in the loop.

Human-reviewed·Agent-assisted·Source-tracked

Loading…

Still in the Loop

A source-tracked, human-reviewed AI newsroom. Agents assist; a human journalist edits and approves. Every story aims to show what changed, why it matters, where it came from, and how much human review it received.

Still in the Loop is new and evolving — we're building it in the open. Spot an error or have a suggestion? stillintheloopai@gmail.com

About & methodology News Feed Contact X AdminEdited by Trevor Brown · Bloomington, IN

Models

DeepSeek-V3

DeepSeek mixture-of-experts language model released in December 2024.

Draftmodelopen-weightsmixture-of-experts

Why it matters

DeepSeek-V3 made model efficiency and open-weight frontier competition a mainstream strategic issue.

Facts

Each fact carries its own source note. Draft and source-check labels mean the item should not be treated as a finished reference entry.

Verifiedhigh confidence
DeepSeek described DeepSeek-V3 as a mixture-of-experts language model with model checkpoints available through its linked repository.[1]
Verifiedhigh confidence
DeepSeek introduced DeepSeek-V3 on December 26, 2024.[1]

Related

companyDeepSeek
conceptOpen weights
conceptAI infrastructure

Source Notes

[1]PrimaryVerified
DeepSeek API Docs: Introducing DeepSeek-V3
Published 2024-12-26. Retrieved 2026-06-07.
Official DeepSeek announcement linking to the DeepSeek-V3 model and technical report.