Open frontier models · San Francisco, US · founded 2024
Deep Cogito: who founded it, who funds it
Deep Cogito is a San Francisco open-weight AI lab founded in 2024 by Drishan Arora and Dhruv Malrana, who came from DeepMind. What it is building: open-weight hybrid reasoning models built by iterated distillation and amplification; post-training for self-improvement. It has raised $56M in disclosed rounds, most recently a Series A of $43M in August 2026, led by TQ Ventures. It releases open-weight models.
- Founders
- Drishan Arora (ex-Google DeepMind), Dhruv Malrana
- Founders from
- ex-Google DeepMind
- Location
- San Francisco, US (North America)
- Valuation
- ·
- Raised
- $56M disclosed
- Investors
- TQ Ventures lead
- What ships
- Open model
Rounds
| Date | Round | Amount | Valuation | Lead | Source |
|---|---|---|---|---|---|
| Aug 26, 2026 | Series A | $43M | · | TQ Ventures | qz.com |
| Jul 15, 2025 | Seed | $13M | · | · | techcrunch.com |
Sources
- WSJ wsj.com
News
- Sep 9, 2026
-
Sep 9, 2026
+7 more coverage
- Quartz Deep Cogito raises $43M Series A for AI self-improvement research
- Pulse 2.0 Deep Cogito Raises $43 Million Series A As Post-Training Platform Targets Self-Improving …
- FinSMEs Deep Cogito Raises $43M in Series A Funding
- citybiz Deep Cogito Raises $43M to Scale Enterprise AI Model Training
- Unite.AI Deep Cogito Raises $43M Series A to Build the Post-Training Engine for Self-Improving AI
- Dealroom Deep Cogito raises $43M to help companies own their AI models
- TradingView Deep Cogito Raises $43 Mln In Series A Led By TQ Ventures
-
Aug 26, 2026
+1 more coverage
- SiliconANGLE Deep Cogito raises $43M to develop self-improving AI models
-
Apr 8, 2025
Same research area
- Reflection AI $25B · New York, US
- Applied Compute $3.25B · San Francisco, US
- Nous Research $1.5B · New York, US
- Essential AI $1B · San Francisco, US
- Prime Intellect $1B · San Francisco, US
- DeepSeek $52B · Hangzhou, CN