Stories
Page 3 of 6 · newest source date first.
Latest stories
- Desktop
Nvidia ships Nemotron 3.5 Lightning and NeMo Switchyard, citing 28% lower Devin Desktop costs
Posts relaying Nvidia's launch say Cognition used the new open-source router in Devin Desktop at 28% lower mean cost. Cognition has not published its own figures.
- Cognition
Who profits when your agent is inefficient? Cognition's president keeps asking
Russell Kaplan made the same argument three times in two weeks — on misaligned incentives, on why SWE-bench got too easy, and on model routing. Together they read as one thesis: per-token economics reward waste, and Cognition is betting its product line on the alternative.
- Cognition
"Infinite AI interns": what Scott Wu's framing actually asks of your team
The Cognition CEO's summer media run — a hands-on Lenny's Newsletter session, a RAISE Summit keynote, and a $26B-valuation interview — sells Devin as capacity, not headcount replacement. The workflow he demos makes specific demands: scoped tasks, public channels, and humans who still own review.
- Cognition
One year after the Cognition–Windsurf deal: what actually shipped
Cognition's anniversary note records a 72-hour acquisition, three homegrown models, a unified Desktop, and revenue run rate from $73M to $500M+. Here is what the year actually produced — grounded, with the receipts.
- CLI
Cognition partners with LTM to put Devin inside managed security remediation
LTM is embedding Devin Security Swarm into BlueVerse RightLogic, a managed service that ingests scanner findings and ships remediation PRs — targeting 80% of an enterprise's CVE backlog. First of five joint offerings, starting with banking.
- Cognition
Cognition: Solving Open Math Problems with 100+ Parallel Devins
Here's a summary of the concrete claims, demos, limitations, and source-worthy details about Devin and Cognition from the video: Devin Product Claims & Demos: Automated Problem So…
- Cognition
Cognition: Welcoming The Interaction Company
Cognition welcomes The Interaction Company (Poke)—a proactive personal agent native to Apple Messages—to deepen always-on cloud agent collaboration with Devin.
- Cognition
Cognition: Signs MOU with U.S. Department of Energy to Join The Genesis Mission
Cognition signed an MOU with the DOE to join the Genesis Mission—contributing Devin for scientific software security, legacy modernization, researcher capacity, and cloud migration.
- Cognition
Cognition: Welcoming TierZero to Cognition
Cognition welcomes Anhang, Yun, and the TierZero team to deepen Devin automations for incidents and production health—not just writing code.
- Cognition
When agents open 200 PRs, Devin Review is the real rate limit
Cognition says code review is the new bottleneck. Developers Digest hit it after a parallel fan-out day. Here’s what Devin Review actually does—and what it still leaves to humans.
- Cognition
Windsurf is Devin Desktop now — and Cascade’s July runway is the real deadline
What Cognition’s June 2 rename actually changed: Agent Command Center first, Devin Local as Cascade’s successor path, ACP for third-party agents, and CLI /handoff when the laptop is not enough.
- Cognition
One year of Cognition + Windsurf: building Devin Desktop together
Cognition’s anniversary note on a year of Windsurf → Devin Desktop — the same story that now shows as Command Center–first Desktop.
- Cognition
Cloud, Local, and Cascade: the three ways Devin actually works
A clear, cited field guide to Cognition’s cloud sessions, Devin Local / CLI, and Cascade — for builders who pair and ship through CI.
- Cloud
Fable 5 costs more per token—but Fusion made Devin’s bill cheaper than Opus
On FrontierCode 1.1 with Fusion sidekicks, Fable 5 led runs scored higher and cost less per run than Opus 4.8—because the lead model’s management style dominates token price.
- Cognition
Devin clears FedRAMP High In-Process for federal engineering teams
Cognition’s full platform is FedRAMP Class D (High) In-Process and listed on the Marketplace—extending High beyond Desktop to Devin Cloud for defense and civilian agencies.
- Cognition
Cognition Measures Trust in Open-Source Models
Cognition evaluates trustworthiness of open-source-derived models, ensuring neutrality and security in autonomous software engineers
- Cognition
Cognition Unveils SWE-1.7 Model
Cognition launches SWE-1.7, a frontier-level intelligence model at a lower cost, advancing the cost-performance Pareto curve
- Cognition
Cognition Refines FrontierCode 1.1
Cognition releases FrontierCode 1.1 with improved methodology for fair internet use and grading criteria
- Cognition
Cognition Introduces Devin Security Remediation
Cognition launches the Devin Security Vulnerability Remediation Program to help organizations clear vulnerability backlogs
- Cognition
Cognition: Introducing FrontierCode
Today’s coding benchmarks have established that models can write correct code, but the question we should really be asking is: can models actually write good code?
- Cognition
Cognition: Estimating the Productivity of an Autonomous AI Software Engineer
Engineering leaders want to know how much value AI is actually providing. We built a system that measures the number of human-equivalent hours of Devin's work.
- Cognition
Cognition: AI should earn its keep: Introducing the AI Productivity Guarantee
Introducing the AI Productivity Guarantee for enterprise customers. If Devin delivers less engineering value than you’re paying for, Cognition will fund your usage until it does,…
- Cognition
Cognition: Verifying Agentic Development at Scale
What we’ve learned building end-to-end testing capabilities in Devin’s virtual machine
- Cognition
Cognition: More Devins in More Places
Cognition has raised over $1B at a $26B valuation, led by Lux Capital, General Catalyst, and 8VC.