Cognition AI Launches Devin 2.0: AI Coding Assistant Gets Major Upgrade and Price Cut

Devin, the viral coding AI agent, gets a new pay-as-you-go plan | TechCrunch

Cognition, the startup behind the viral AI programming tool Devin, has introduced a new low-cost plan to incentivize sign-ups. When Cognition released Devin last year, the tool quickly blew up on social media for its ability to perform certain software development tasks autonomously. It quickly became apparent that Devin struggled with more complex coding work. Nevertheless, the tool garnered praise from AI founders including Perplexity CEO Aravind Srinivas and substantially raised Cognition's profile. Devin launched in general availability for teams starting at the eye-watering price of $500 per month. On Thursday -- conspicuously weeks after the company reportedly raised hundreds of millions of dollars in fresh capital -- Cognition introduced an entry-level option that costs $20, then transitions to pay-as-you-go. The pay-as-you-go plan could end up being quite costly, depending on how one uses Devin. $20 nets you around 9 ACUs, Cognition's jargon for computing credits. (ACUs cost $2.25 on the $20 plan, a hike from the $2 they cost on the $500-per-month subscription.) Cognition says that 15 minutes of "active Devin work" is equivalent to about 1 ACU. 9 ACUs only nets around 2.25 hours of work, by that metric -- not much when you're dealing with massive codebases. But Cognition claims that Devin today -- Devin 2.0 -- is much improved compared to the December release. Similar to GitHub's Copilot tool, Devin can now help generate plans for coding projects, as well as answer questions about code with citations and create "wikis" for code with documentation. Silas Alberti, a member of Devin's development team, also told TechCrunch that Devin now "gets twice as much work done as before." Those claims are best taken with a grain of salt. Even the best code-generating AI today tends to introduce security vulnerabilities and bugs, studies have found, owing to weaknesses in areas like the ability to understand programming logic. One recent evaluation of Devin found that it completed just three out of 20 tasks successfully.

Devin 2.0 is here: Cognition slashes price of AI software engineer to $20 per month from $500

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More The Founders Fund-backed San Francisco startup Cognition AI (also known as Cognition Labs) made a name for itself out of nowhere in early 2024 with the release of Devin, its AI-powered software engineer that could work alongside human developers and carry out tasks autonomously through natural language instructions given to them by a human dev through a prompt window or even separate third-party communication app Slack. But the AI dev landscape has advanced rapidly since then, with many other companies offering similar functionality and autonomous or semi-autonomous coding agents, including GitHub Copilot, AWS Developer Q, Codeium's Windsurf, and Cursor. Cognition itself has since its inception leveraged other models, namely OpenAI's GPT-4 and GPT-4o series, to power Devin. Now today, Cognition is hitting back with Devin 2.0, an updated version of its agent-native software development platform. It's unclear what foundation model is powering this version. The new release is now generally available and introduces a range of features aimed at making collaboration between developers and Devin's autonomous agent smoother and more productive. Furthermore, in an age of economic uncertainty and certain cost increases thanks to the new Trump tariffs, Cognition is also providing a welcome relief with a major price drop: Devin 2.0 is available starting at $20 per month minimum ($2.25 per "Agent Compute Unit", how Cognition measures the compute resources required to run Devin), whereas before, previous versions of the software started around $500 per month. Devin 2.0 builds on Cognition Labs' earlier efforts to streamline software development by allowing users to work alongside autonomous agents. The latest version introduces an interactive, cloud-based IDE environment designed to allow users to spin up multiple Devins in parallel, effectively handling numerous tasks at once. Each Devin can work autonomously, with users having the option to step in at any point to review, edit, or guide progress. Interactive planning and task scoping A key addition in Devin 2.0 is the introduction of Interactive Planning. This feature allows developers to begin with broad or incomplete ideas and collaborate with Devin to scope out a detailed task plan. Within seconds of starting a session, Devin automatically analyzes the codebase, identifies relevant files, and proposes an initial plan, even without specific guidance or instructions from the human user. Users can then review and adjust this plan to ensure alignment before allowing Devin to proceed with execution. Exploring your enterprise's codebase Devin 2.0 also introduces Devin Search, a tool designed to help users understand and navigate their codebases more effectively. The search function enables developers to ask specific questions about their code, receiving detailed responses that cite relevant code snippets. For more complex queries that require deeper exploration, users can activate Deep Mode. Additionally, the new release includes Devin Wiki, a feature that automatically indexes repositories every few hours. Devin Wiki generates comprehensive documentation that includes architecture diagrams, source links, and other relevant details, offering developers an organized and continuously updated reference. Efficiency gains and developer control Beyond new features, Cognition Labs reports that Devin 2.0 delivers improved efficiency. According to the company, the latest version completes over 83% more junior-level development tasks per Agent Compute Unit (ACU) compared to its predecessor, based on internal benchmarks. Beta users reportedly observed similar performance gains during testing. Users can interact with Devin 2.0 through a VSCode-inspired interface that allows for reviewing and editing Devin's work, as well as running tests directly within the platform's environment. This flexibility supports both hands-on and hands-off workflows, depending on user preference. A boost from Devin 1.2 In early 2025, the company released Devin 1.2, which included enhancements focused on in-context reasoning and voice command integration. These improvements enabled Devin to better analyze code repositories, recognize patterns, and reuse existing code when appropriate. Users could also issue instructions via voice messages in Slack, streamlining how they interacted with the agent. Devin 1.2 also introduced features geared toward enterprise environments, such as machine snapshots to simplify login workflows and centralized admin controls for managing multiple Devin workspaces. Alongside these functional upgrades, Cognition Labs shifted to a usage-based billing model, allowing customers to pay for additional capacity beyond their subscription limits. How Devin stacks up now to other AI coding agents and platforms While Devin's early releases positioned the platform as an innovative solution for accelerating development workflows, early user feedback highlighted growing pains. Researchers and testers noted that the agent sometimes struggled with overly complex code, unnecessary abstractions, and inconsistent task performance. Nevertheless, Devin attracted interest from enterprise customers seeking to incorporate autonomous coding agents into their software development processes. The new features and capabilities of Devin 2.0 -- and a much lower entry price -- should be received warmly by devs, and may further uptake of Cognition's platform by them, even encouraging users to defect from other rival coding tools. But with GitHub Copilot, Codeium's Windsurf, and Amazon Q Developer among others all offering free versions of their AI coding assistants, Devin 2.0 faces an increasingly tough set of competitors in a white hot market.

Cognition AI launches revamped coding assistant Devin 2.0 with much lower starting price - SiliconANGLE

Cognition AI launches revamped coding assistant Devin 2.0 with much lower starting price Cognition AI Inc., creator of the generative artificial intelligence-powered coding bot Devin, today announced a new version of that tool with more powerful capabilities, and it's available at a much lower cost than before. It was just over a year ago that Cognition AI took the coding world by storm with the release of Devin, which far surpassed AI coding assistants such as GitHub Inc.'s Copilot. Unlike existing coding bots, which could only assist human developers, Devin was touted as being able to autonomously create the codebase of entire applications, including new AI models. Devin went viral, with its ability to transform a natural language prompt into a full-blown application, writing its own code, testing and reporting on its progress in real-time. Some analysts said it marked the dawn of a new era where software creates software, with the role of human programmers limited to supervising its work. However, it soon became apparent that Devin wasn't all that it cracked up to be, struggling with more complex coding tasks, and reports of buggy code that didn't always work as it was supposed to. Devin was also extremely expensive, priced at $500 per user per month. Moreover, many rival offerings with similar functionality emerged in its wake, with GitHub's Copilot getting more autonomy, alongside offerings like Amazon Web Services Inc.'s Developer Q, Magic AI Inc,'s LTM-2-mini and Codeium Inc.'s Windsurf. Due to those realities, Devin's adoption stalled, but Cognition AI is now hitting back. With the launch of Devin 2.0 in beta today, the company has rolled out a new pricing plan that starts at just $20 per month, transitioning to a pay-as-you-go model afterwards. It also introduces a range of capabilities that aim to streamline collaboration between developers and Devin and increase productivity. Whether or not the new pricing plan substantially lowers the overall costs remains to be seen, though. For the starting price of $20, customers gain access to 9 ACUs, which is the company's name for credits. That means users are paying $2.25 per ACU, as compared to the $2 they were paying under the original $500 per month subscription. According to Cognition AI, one ACU equates to 15 minutes of "active work", so that translates to a cost of $11 per hour - not so different from what a freelance human developer on a platform such as Upwork might charge. Of course, the advantage of Devin 2.0 is that it can work much more quickly than any human programmer can, and with today's update it is far more capable than the original version. For instance, it's now able to help create plans for coding projects, answer questions about its code with citations, and even create a "wiki" for the code it generates. One major change is the introduction of a cloud-hosted Integrated Development Environment or IDE, which enables multiple Devins to be operated at once, with each one handling a different task. The idea is that the human user simply sits there and oversees what the Devins are doing, stepping in when required. According to the startup, this feature can help companies to better organize their programming teams, giving them greater flexibility. Another intriguing feature is Devin 2.0's new Interactive Planning tool, which allows human users to outline a nebulous idea. Devin then takes over, fleshing out hazy concepts to create a detailed plan of action that allows it to be brought to life. Devin's new analytics capabilities should come in handy too. Among other things, it can analyze an entire codebase, identify the relevant files within it, and then propose action plans to correct or improve the code - all without human input. In addition, there's a new Devin Search feature, which allows for deeper exploration of code, while the Devin Wiki tool will help teams to ensure each codebase is thoroughly documented and kept up to date. The idea is that these features can help teams to keep a lid on the chaos that typically emerges with large-scale software projects. The big question is whether or not these changes will help Devin and Cognition AI to regain their viral mojo in an area that has seen significant advances over the intervening months. Although Devin positions itself as a platform for accelerating developer workflows, the reality has been somewhat different. There have been numerous reports of Devin struggling to get to grips with more complex codebases, as well as inconsistent performance and the inclusion of unnecessary abstractions. Of course, Cognition AI will argue these are standard growing pains for a new technology, and with its lower entry price and improved functionality, it will be hopeful of encouraging further uptake of Devin in the enterprise. But we can expect its rivals to keep pace with these developments and introduce their own innovations, too.

Devin 2.0 Released; New Plan Offers 96% Price Cut

Cognition AI also announced additional features for Devin, including Interactive Planning, Devin Search, and Devin Wiki. US-based Cognition AI has announced an update to Devin, which was dubbed the 'world's first AI software engineer' upon its release last year. Now, Devin 2.0, its updated version, offers a new agent-native IDE experience. Devin 2.0 supports multiple parallel instances, each with an interactive cloud-based IDE. Additionally, the latest update allows developers to take control while providing collaborative and fully automated approaches. Furthermore, it enables developers to refine code and run tests within the IDE. Cognition AI also announced additional features for Devin, including Interactive Planning, Devin Search, and Devin Wiki. Interactive Planning offers a comprehensive, editable task plan at the beginning of each session, derived from an analysis of the codebase. Devin Search enables developers to query their codebase with referenced answers and includes a Deep Mode for more intricate queries. Devin Wiki automatically refreshes repository documentation, encompassing architecture diagrams and source links. All these features can be accessed on Devin's web-based platform. Moreover, Devin now offers a flexible, entry-level pay-as-you-go plan starting at just $20. This plan includes a specified number of Agent Compute Units (ACUs), with any additional ACU priced at $2.25 once the initial allocation ($20) is consumed. It features unlimited access to seats and requires no monthly commitment. This represents a substantial decrease from the $500 monthly plan for teams, which was announced earlier. ACUs evaluate Devin's workload by taking into account the number and complexity of actions carried out, as well as virtual machine and network usage. The company also mentioned that each ACU now achieves nearly double the output according to their internal benchmark for junior developer tasks. Devin 2.0 accomplishes over 83% more tasks per ACU, demonstrating a rise in efficiency that has also been observed among their beta users. When Devin was announced last year, developers testing the tool were disappointed by several shortcomings. However, with updates, the tool appears to have improved, as some developers have increasingly praised its capabilities. Sahil Lavingia, CEO of Gumroad, revealed that Devin completes a large number of PRs (pull requests) in his organisation. In February, he said that Devin is the most 'productive' engineer in his company, measured by merged PRs. The co-founder of Cognition AI, Scott W, is a three-time gold medal winner at the International Olympiad of Informatics and held the third position in 2021's edition of the prestigious Google Code Jam competition.