CURATED TIMELINE · EDITORIAL EDITION

DeepSeek Before the Breakout

The January 2025 sensation did not appear overnight. Trace the open-weight code models, sparse experts, GRPO, MTP, V2 efficiency work, R1 preview, V3, the free app, and the App Store shock — while separating reported training-run cost from total R&D and open weights from full open source.

Timeline overview

At a glance

Hover for the event briefScroll horizontally

2023.112025.01

11.02DeepSeek Coder 01.11DeepSeekMoE 02.05DeepSeekMath / GRPO 05.07DeepSeek-V2 11.20R1-Lite-Preview 12.26DeepSeek-V3 01.27DeepSeek goes mainstream

10 nodes2023.11—2025.01 · Editorially curated · Updated 2026-07-15

Editorial thread

The timeline

Reviewed event briefs and original editorial context, ordered to show how the story changed over time.

01
2025-01-20Open-source release
DeepSeek releases the open-weight reasoning model R1
DeepSeek publishes R1, R1-Zero, six distilled models, and its reinforcement-learning recipe, pushing open reasoning models into the frontier conversation.
Editorial context
R1 turned the research program into a complete launch. DeepSeek released R1-Zero, R1, six distilled checkpoints, a paper, web access, and an API at once. Downloadable weights and permissive reuse created a sharp contrast with o1's product-and-API-only access.

Benchmark parity should be read task by task, and low API prices do not by themselves prove equally low serving cost or the absence of subsidy. The verifiable shift was access: developers could inspect, run, adapt, and distill the model.
Read event
02
2025-01-27Market shift
DeepSeek tops the U.S. App Store and jolts AI markets
DeepSeek passes ChatGPT on the U.S. App Store's free chart as Nvidia loses roughly $593 billion in market value during an AI-sector selloff.
Editorial context
This was the translation layer from technical debate to mass culture. A U.S. App Store number-one badge made “open, capable, cheap” understandable without reading a benchmark table. Nvidia's record market-value decline on the same day then pulled the story into finance and politics.

The ranking measured download momentum, not total active users or retention. Roughly $593 billion was a market repricing of future expectations, not money directly destroyed by DeepSeek.
Read event

DeepSeek releases the open-weight reasoning model R1

DeepSeek tops the U.S. App Store and jolts AI markets