DEPLOYDatabase

Company

Cognition

cognition.ai

Cognition is an applied AI lab behind Devin, an AI software engineer.

Track Cognition
Status
Private

Overview

Cognition is an applied AI lab behind Devin, an AI software engineer. Runs training, reinforcement learning, and production inference on CoreWeave; first customer reported running production workloads on NVIDIA Vera Rubin NVL72.

Verified record

Verified deployments
None on file
Active incidents
None on file

Key facts

First Vera Rubin NVL72 production customer on CoreWeave (company-stated, Sep 2026)

Company-stated (company-stated; also NVIDIA Blog): Cognition (applied AI lab behind Devin) is the first customer anywhere running production workloads on NVIDIA Vera Rubin NVL72, on CoreWeave Cloud. Cognition runs training, RL, and production inference for Devin on CoreWeave; cluster stood up early September with workloads running within days of rack handover per CoreWeave. Cap: CoreWeave/NVIDIA company press.

SWE-2 up to 4.8x token throughput vs GB200 on Vera Rubin (company-stated, Sep 2026)

Company-stated (company-stated; also NVIDIA Blog): Cognition reports up to 4.8x total token throughput for SWE-2 inference vs GB200 NVL72 at matched interactivity framing, and 3.8x output-token throughput for RL workloads. Cap: company-stated customer benchmark.

Data & sources

Web sources

3

3 sources backing this record.View all →

Safety record

No incidents on record for Cognition.

Only active incidents are counted. Retracted incidents are excluded from this summary but remain reachable at their canonical URLs.

Full safety record: incidents, sourcing, and exposure data →