Mark RadarMARK RADAR
About
EN
Sign in
Topic File

DeepSeek

DeepSeek is a Chinese AI model developer known for its open-source, low-cost large language models and reasoning technology. In addition to plans for a Shanghai listing and the completion of a major funding round, the company is developing its own AI inference chips and the open-source DSpark acceleration framework. Meanwhile, China is tightening regulation and the U.S. Congress is investigating the risks of companies adopting Chinese models. As Chinese models rapidly gain ground on OpenRouter and in global token usage, DeepSeek warrants close attention because its development has implications for AI costs, the open-source ecosystem, chip demand and U.S.-China technology competition.

47 core events · 27 peripheral · Tracking since 2026-02-24 · Last active 2026-09-01 · ai 47

Key Moments

7 selected

The full history is split into 7 equal periods by event count, taking the most-covered story from each. Coverage rises and falls with the news cycle, so sampling period by period keeps the most recent events from taking everything.

  1. 2026-08-14 DeepSeek Open-Sources Plugin-Based Harness for AI Agents 4 reports
  2. 2026-08-05 DeepSeek Restarts $8 Billion Funding Round at 500 Billion Yuan Valuation 3 reports
  3. 2026-08-01 DeepSeek Upgrades V4-Flash to Boost Agentic Coding 3 reports
  4. 2026-07-24 DeepSeek Teams With Huawei to Challenge Nvidia’s CUDA Moat 2 reports
  5. 3 more key moments Sign up to see the full context

Event Timeline

47 core events
27 peripheral mentions — events mainly about something else

You have used up your follow slots

The free plan caps how many topics you can follow. More slots and a daily digest are being planned — tell us what you think, and we'll let you know the moment it opens.

If there were a paid plan, roughly how much would you pay?

Nothing about the plan or pricing is settled yet; this answer only helps decide whether and how to build it.

Mark Radar|MARK RADAR
All times are in Taipei time (GMT+8)