Devin
An autonomous AI agent associated with Cognition. Here it is described as participating in long-running agent threads and workflow automation with trybasis.
Key Highlights
- Devin is presented here as Cognition’s autonomous AI agent spanning coding, workflow automation, and long-running task execution.
- Newsletter coverage highlights key product themes: lower cost, stronger integrations, persistent memory, and support for external model ecosystems.
- The trybasis mention is especially notable for AI PMs because it describes 20-day agent threads in a real workflow automation context.
- Enterprise relevance is reinforced by integrations with Microsoft Teams and Microsoft 365, plus operational usage tied to GitHub, Sentry, and bug investigation.
- For AI PMs, Devin is a useful case study in measuring agent value through task duration reduction, alert triage impact, and workflow economics.
At a glance
- Category
- AI Agent
Devin
Overview
Devin is an autonomous AI tool associated with Cognition, positioned as an agent for software and operational workflows rather than just a chat interface. In the newsletter coverage here, Devin appears in several forms and surfaces—including Devin Desktop, CLI, cloud-based usage, and integrations—and is described as participating in long-running agent threads, workflow automation, bug investigation, review workflows, and enterprise collaboration environments.For AI Product Managers, Devin matters because it represents a shift from single-turn copilots to persistent, multi-step agents that can operate across tools, sessions, and teams. The mentions here show Devin expanding on three important dimensions: deeper enterprise integrations, lower operating cost, and stronger memory/workflow capabilities such as 20-day agent threads and cross-session memory graphs. Together, these signal a practical move toward deployable AI agents that can own parts of engineering and ops workflows.
Key Developments
- 2026-09-24: Cognition announced native Microsoft Teams support for Devin plus a first-party Microsoft 365 integration connecting to mail, calendar, files, chats, and more.
- 2026-09-29: Cognition said Devin became 30–40% cheaper in Fusion and Normal, 15–20% cheaper in Ultra, and up to 70% cheaper in Devin Review, while also improving capabilities.
- 2026-10-01: Cognition shared a case study from Crosbylegal: Devin handled first-response investigation for 20–40 bugs and alerts per day, reduced Sentry noise by 50%, and helped keep on-call to one engineer as the legal team grew.
- 2026-10-02: Devin was listed among the launch partners for Sign in with ChatGPT, enabling users to spend their own ChatGPT token balance instead of developer-paid tokens.
- 2026-10-06: Cognition introduced Dreaming for Devin, a capability that builds a cross-session memory graph of user preferences and refines it overnight by removing stale records and discovering latent information.
- 2026-10-10: Cognition announced users could connect a personal ChatGPT Go, Plus, or Pro plan to Devin, with GPT model usage drawing from the connected plan quota.
- 2026-10-23: Cognition open sourced a security-focused agent plugin that can be used with Devin or installed with any agent to help secure systems against recent cybersecurity threats.
- 2026-10-25: AWS’s agent-toolkit-for-aws highlighted support for Devin alongside Claude Code, Codex, Cursor, Kiro, and fx, signaling Devin’s place in the broader coding-agent ecosystem.
- 2026-11-11: Cognition announced that Google’s Gemini 4 Carbon became available in Devin Desktop and CLI.
- 2026-11-12: Cognition shared how trybasis works with Devin, including 20-day agent threads; the example framed Devin as part of long-running workflow automation that helped reduce 20-hour accounting tasks to 30 minutes.
Relevance to AI PMs
1. Evaluate agent durability, not just model quality. Devin’s 20-day threads and Dreaming memory features are useful reference points when scoping products that need persistent context across sessions, owners, and workflows.2. Design integrations around where work already happens. The Microsoft Teams, Microsoft 365, Slack-adjacent ecosystem, GitHub, Linear, Sentry, and ChatGPT-related mentions show how agent adoption often depends on fitting into existing enterprise systems rather than introducing a standalone interface.
3. Model ROI in operational terms. The case studies and pricing updates provide examples of measurable outcomes AI PMs can track: alert reduction, lower review cost, fewer on-call burdens, faster workflow completion, and better economics for long-running agent tasks.
Related
- Cognition: The company most directly associated with Devin and the source of the product announcements listed here.
- trybasis: A workflow automation context where Devin was described as participating in 20-day agent threads.
- Claude Code, Codex, Cursor, Kiro, fx, Windsurf: Related coding-agent or developer-tooling products that frame Devin within a competitive and interoperable agent landscape.
- GitHub, Linear, Sentry, Slack, Microsoft Teams, Microsoft 365: Workflow and collaboration systems that connect to Devin’s practical usage in engineering and operations.
- ChatGPT, Sign in with ChatGPT, OpenAI, GPT-55, GPT-6 Astra, Gemini 4 Carbon: Model and identity ecosystem references showing Devin’s expanding compatibility with external model providers and user-owned plan/token flows.
- Dreaming, Agent Memory Repo: Memory-related concepts and initiatives connected to Devin’s cross-session personalization and agent-state management.
- security-swarm, bugbot, auto-triage, github-stacked-prs: Adjacent agent workflows and plugins relevant to software operations, code review, and security automation.
Frequently asked questions
What is Devin?
An autonomous AI agent associated with Cognition. Here it is described as participating in long-running agent threads and workflow automation with trybasis.
What should AI product managers know about Devin?
Devin is presented here as Cognition’s autonomous AI agent spanning coding, workflow automation, and long-running task execution. Newsletter coverage highlights key product themes: lower cost, stronger integrations, persistent memory, and support for external model ecosystems. The trybasis mention is especially notable for AI PMs because it describes 20-day agent threads in a real workflow automation context.
Explore this topic
Newsletter Mentions (44)
“Cognition shared how @trybasis works with Devin, including 20-day agent threads.”
#7 𝕏 Cognition shared how @trybasis works with Devin, including 20-day agent threads. @trybasis builds AI agents that the source says cut 20-hour accounting tasks down to 30 minutes.
“Cognition announced that Google’s Gemini 4 Carbon is available in Devin Desktop and CLI.”
#17 𝕏 Cognition announced that Google’s Gemini 4 Carbon is available in Devin Desktop and CLI. It scores 40.0% on FrontierCode 2.0, up from Gemini 4 Argon’s 26.3%, making it the best-performing Google model on the benchmark.
“It supports Claude Code, Codex, Cursor, Devin, Kiro, and fx, and has 3,249 GitHub stars with 13 daily trending gains.”
#2 🐙 aws/agent-toolkit-for-aws aws/agent-toolkit-for-aws - **aws/agent-toolkit-for-aws** is AWS’s official collection of MCP servers, reusable skills, and plugins for helping AI coding agents build, deploy, and manage AWS applications. It supports Claude Code, Codex, Cursor, Devin, Kiro, and fx, and has 3,249 GitHub stars with 13 daily trending gains. Worth trying if you need agent access to AWS APIs, sandboxed scripts, and documentation search.
“#19 𝕏 Cognition open sourced an agent plugin designed to help secure systems against recent cybersecurity threats.”
Cognition open sourced an agent plugin designed to help secure systems against recent cybersecurity threats. It can be installed with any agent or used in Devin. Also covered by: @Cognition
“Cognition announced that users can connect a personal ChatGPT Go, Plus, or Pro plan to Devin, with all GPT model usage drawing from the connected plan’s quota.”
Cognition announced that users can connect a personal ChatGPT Go, Plus, or Pro plan to Devin, with all GPT model usage drawing from the connected plan’s quota.
“Cognition introduced Dreaming, a capability that lets Devin build a cross-session memory graph of how users like to work and improve it at night by removing stale records and discovering latent information.”
#1 𝕏 Cognition introduced Dreaming: Devin refines cross-session memory overnight. Cognition introduced Dreaming, a capability that lets Devin build a cross-session memory graph of how users like to work and improve it at night by removing stale records and discovering latent information. Cognition is also creating an open-source standard called Agent Memory Repo. Also covered by: @Harrison Chase
“Dots costs $100 per month, while Sign in with ChatGPT launched with 16 partners, including Devin, Notion, Vercel, and OpenClaw.”
#10 ▶️ The one OpenAI announcement that can actually make you money... Fireship Fireship recapped OpenAI’s Dev Day 2026 announcements: GPT-6 Astra-powered Dots personal agents, the Decisions API for fixed-option classification with Luna, and Sign in with ChatGPT, which lets users spend their own ChatGPT token balance instead of developers’ tokens. Dots costs $100 per month, while Sign in with ChatGPT launched with 16 partners, including Devin, Notion, Vercel, and OpenClaw.
“Cognition shared that @crosbylegal, with 12 engineers supporting 50+ attorneys, put Devin on first response to investigate 20–40 bugs and alerts a day. Devin cut Sentry noise by 50% and kept on-call to one engineer as the legal team grew 5x.”
Cognition shared that @crosbylegal, with 12 engineers supporting 50+ attorneys, put Devin on first response to investigate 20–40 bugs and alerts a day. Devin cut Sentry noise by 50% and kept on-call to one engineer as the legal team grew 5x.
“Cognition announced updates making Devin 30–40% cheaper in Fusion and Normal, 15–20% cheaper in Ultra, and up to 70% cheaper in Devin Review.”
GenAI PM Daily - Anthropic announces faster, lower-cost Claude Sonnet 5.5 September 29, 2026 GenAI PM Daily 🎧 Listen to this brief 3 min listen Today's top 20 insights for PM Builders, ranked by relevance from Blogs, X, YouTube, and LinkedIn. Anthropic announces faster, lower-cost Claude Sonnet 5.5 #1 📝 Anthropic News Introducing Claude Sonnet 5.5 - Anthropic announced Claude Sonnet 5.5 on September 28, 2026—the second model in the Claude 5.5 family—which it says runs 30%+ faster and costs up to 30% less for most work. Positioned as a faster, lower-cost complement to Claude Opus 5.5, it is strongest at well-scoped everyday tasks, bug fixes, and creating polished documents. Also covered by: @Peter Yang , @Cognition , @v0 , @Cursor , @Boris Cherny , @Mike Krieger , @Mike Krieger , @Anthropic , @Claude – Anthropic , @Claude – Anthropic , @There's An AI For That #2 𝕏 Cognition announced updates making Devin 30–40% cheaper in Fusion and Normal, 15–20% cheaper in Ultra, and up to 70% cheaper in Devin Review. Cognition also said the updates improve capabilities, with Devin Fusion now ranking first on FrontierCode 1.1 Extended.
“Cognition announced two updates to Devin for Microsoft users: native Microsoft Teams support and a first-party Microsoft 365 integration connecting to mail, calendar files, chats, and more.”
#11 𝕏 Cognition announced two updates to Devin for Microsoft users: native Microsoft Teams support and a first-party Microsoft 365 integration connecting to mail, calendar files, chats, and more. Also covered by: @Cognition
Related
The AI company referenced in relation to research on chain-of-thought. It is relevant as the source of the research being discussed.
A coding assistant used for execution and implementation work in the tutorial referenced by Peter Yang. It is presented as part of a workflow paired with different models for different tasks.
Cursor is an AI coding tool that introduced Team Projects in beta for shared agent context across teammates. The update emphasizes coordination, conflict handling, and team-level task unblocking.
An AI coding tool used to build Mural and referenced as a model for fast-moving product workflows. It is also mentioned as an environment where Garry Tan runs evals.
AI company that builds Devin and related developer tools for agentic software work.
A conversational AI tool referenced as part of a project-only-memory pattern for handling sensitive information. The newsletter uses it as an example of a siloed memory setup for medical records.
A partner company listed in the Sign in with ChatGPT launch set. The newsletter provides no additional context beyond the integration mention.
A workplace messaging platform referenced as a target surface for generative UI in conversational agents.
Linear is a product-management and issue-tracking platform that is integrating AI agents into workflow assignment. Here it is highlighted as the place where Amp can be assigned issues directly.
A model used as an automated judge in Claire Vo’s benchmark. It contributes 30% of the scoring alongside her manual evaluation.
Platform provider referenced through its spec-kit toolkit and repository statistics. For PMs, it indicates GitHub’s role in shaping AI coding workflows and agentic software development.
AI product management assistant used for product discussions, including fraud handling and trade-offs. Claire Vo says she spends much of her time on it dealing with token fraud operations.
Ramp is a fintech/spend-management platform used here as the source of firm-level AI spending data. It provides the benchmark for comparing high-spending and median customers.
A model used in the newsletter as a reasoning and execution engine for product experimentation. It is described as generating daily A/B test ideas and implementing winners for a mobile game economy.
An OpenAI model referenced as part of the Daybreak Blue alpha for defensive vulnerability validation. It appears to support security testing and fix verification workflows.
AI product company mentioned both as Elena Verna’s employer and as a product she uses in her pricing and packaging work.
Stay updated on Devin
Get curated AI PM insights delivered daily — covering this and 1,000+ other sources.
Subscribe Free