Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Watch Devin Fusion work: The sidekick model handles fetching information and skimming sources. The frontier model makes a plan for implementation, then delegates well-scoped sub-tasks to the sidekick with its own cached context.

Cognition

171,076 subscribers

25,080 views • 3 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Devin tests its work before you review the PR. You review and approve the test plan. Then get back a screen recording with a visual checklist of step-by-step QA.

Devin tests its work before you review the PR. You review and approve the test plan. Then get back a screen recording with a visual checklist of step-by-step QA.

Cognition

64,992 views • 3 months ago

Devin can manage a team of Devins. Managed Devins run in parallel help break down complex tasks. Each managed session is a full Devin, with its own VM, terminal, browser, and testing infrastructure. The main session coordinates, monitors, and compiles results.

Devin can manage a team of Devins. Managed Devins run in parallel help break down complex tasks. Each managed session is a full Devin, with its own VM, terminal, browser, and testing infrastructure. The main session coordinates, monitors, and compiles results.

Cognition

31,758 views • 5 months ago

Starting today, you can now use your ChatGPT subscription directly in Devin. Just sign in to Devin with your ChatGPT Plus or Pro plan and all OpenAI model usage in Devin will drawn down from your quota. Now live in Devin Cloud, Devin Desktop, and Devin CLI.

Starting today, you can now use your ChatGPT subscription directly in Devin. Just sign in to Devin with your ChatGPT Plus or Pro plan and all OpenAI model usage in Devin will drawn down from your quota. Now live in Devin Cloud, Devin Desktop, and Devin CLI.

Cognition

449,327 views • 14 days ago

Today we're excited to introduce Devin, the first AI software engineer. Devin is the new state-of-the-art on the SWE-Bench coding benchmark, has successfully passed practical engineering interviews from leading AI companies, and has even completed real jobs on Upwork. Devin is an autonomous agent that solves engineering tasks through the use of its own shell, code editor, and web browser. When evaluated on the SWE-Bench benchmark, which asks an AI to resolve GitHub issues found in real-world open-source projects, Devin correctly resolves 13.86% of the issues unassisted, far exceeding the previous state-of-the-art model performance of 1.96% unassisted and 4.80% assisted. Check out what Devin can do in the thread below.

Today we're excited to introduce Devin, the first AI software engineer. Devin is the new state-of-the-art on the SWE-Bench coding benchmark, has successfully passed practical engineering interviews from leading AI companies, and has even completed real jobs on Upwork. Devin is an autonomous agent that solves engineering tasks through the use of its own shell, code editor, and web browser. When evaluated on the SWE-Bench benchmark, which asks an AI to resolve GitHub issues found in real-world open-source projects, Devin correctly resolves 13.86% of the issues unassisted, far exceeding the previous state-of-the-art model performance of 1.96% unassisted and 4.80% assisted. Check out what Devin can do in the thread below.

Cognition

31,482,406 views • 2 years ago

Devin can now manage a team of Devins. Devin will break down large tasks and delegate them to parallel Devins that each run in their own VM. Over time, Devin gets better at breaking down and managing tasks for your codebase. Available now for all users.

Devin can now manage a team of Devins. Devin will break down large tasks and delegate them to parallel Devins that each run in their own VM. Over time, Devin gets better at breaking down and managing tasks for your codebase. Available now for all users.

Cognition

111,653 views • 6 months ago

Building and prototyping games is easier than ever with Sidekick. Our designer built a 3D version of this OpenAI GPT-o1 game demo in <1 hour. Used Autodesk Maya for assets, Unity for engine, @bezi_3d Sidekick for scripting & implementation. Sidekick analyzes your project for real-time context of the asset folder, scene objects, components, and more. This allows it to implement simple mechanics—like movement controls and basic physics—for your characters and objects in seconds. It's the AI tool built by game developers, for game developers.

Building and prototyping games is easier than ever with Sidekick. Our designer built a 3D version of this OpenAI GPT-o1 game demo in <1 hour. Used Autodesk Maya for assets, Unity for engine, @bezi_3d Sidekick for scripting & implementation. Sidekick analyzes your project for real-time context of the asset folder, scene objects, components, and more. This allows it to implement simple mechanics—like movement controls and basic physics—for your characters and objects in seconds. It's the AI tool built by game developers, for game developers.

Bezi

15,295 views • 2 years ago

Sub-agent Model Selection — Different Tasks, Different Models Your main agent runs Qwen3.6-Plus for quality. But not every subtask needs a flagship model. Now sub-agents can use a different model. Create a skill file with model: openai:qwen3.5-plus and the sub-agent runs on that model. Powerful model for the hard parts, fast model for the easy parts. Save tokens without sacrificing quality on what matters.

Sub-agent Model Selection — Different Tasks, Different Models Your main agent runs Qwen3.6-Plus for quality. But not every subtask needs a flagship model. Now sub-agents can use a different model. Create a skill file with model: openai:qwen3.5-plus and the sub-agent runs on that model. Powerful model for the hard parts, fast model for the easy parts. Save tokens without sacrificing quality on what matters.

Qwen

21,333 views • 6 months ago

The phone threshold is even harder than the storage number suggests. A phone exposes only part of its memory to an application, and the model must share that budget with its KV cache, activations, runtime, and the rest of the product. At 3.9 GB, 1-bit Bonsai 27B clears that threshold with room to work, making 27B-class local AI possible on a phone for the first time. (Demo Mode: cached & prefilled image context)

The phone threshold is even harder than the storage number suggests. A phone exposes only part of its memory to an application, and the model must share that budget with its KV cache, activations, runtime, and the rest of the product. At 3.9 GB, 1-bit Bonsai 27B clears that threshold with room to work, making 27B-class local AI possible on a phone for the first time. (Demo Mode: cached & prefilled image context)

PrismML

33,665 views • 3 months ago

Introducing Harvey II: smarter agents from the start. - Featuring Harvey Tenet, our first model trained for legal work - Built around matters and projects - Agents start with the files, context, permissions, and history they need - Assign tasks to lawyers or agents, then track and review the work - Harvey remembers how you work and writes like you

Introducing Harvey II: smarter agents from the start. - Featuring Harvey Tenet, our first model trained for legal work - Built around matters and projects - Agents start with the files, context, permissions, and history they need - Assign tasks to lawyers or agents, then track and review the work - Harvey remembers how you work and writes like you

Harvey

248,354 views • 1 month ago

Progress in open models is keeping Big AI labs up at night, and I'm here for it! We have a brand new open-weight multimodal model optimized for long-horizon tasks. This model is really good at something: it can work on tasks that keep evolving over time. • 280B total parameters, but only 16B active • 512K context window • Understands text, images, and audio • Strong reasoning, coding, and tool use But the best of all: the model learns and adapts to new information! Imagine you start running an agent today to solve a problem, and while it's working, you get new information that changes the initial conditions, or you change your mind. The agents you run today don't have issues with short tasks and goals that don't change, but reality is messy, and that makes it hard for long-horizon agents to succeed. The new dots3-note Preview model introduces TEMPO. TEMPO is a new reinforcement learning technique that lets the model periodically pause and critique its own progress. Basically, from time to time, the agent asks itself: "Am I getting closer to the goal, or am I wasting my time?" The same model switches between actor and critic. The actor works on the problem. The critic looks at the current state, reasons about how much progress it has made, and determines what should happen next. TEMPO gives the model feedback along the way. This is huge for any agent that can work on long-horizon tasks without wasting its time.

Santiago

80,792 views • 1 month ago

If your AI can't call a function or hit an API, it's not really an agent (it's a chatbot). Vision Agents makes it simple to connect MCP servers and function calling so your model can interact with its surroundings: running code, fetching data, and automating tasks. Step-by-step implementation from Stefan in thread ⬇

If your AI can't call a function or hit an API, it's not really an agent (it's a chatbot). Vision Agents makes it simple to connect MCP servers and function calling so your model can interact with its surroundings: running code, fetching data, and automating tasks. Step-by-step implementation from Stefan in thread ⬇

Stream

31,306 views • 11 months ago

Devin is generally available today! Just tag Devin to fix frontend bugs, create first-draft PRs for backlog tasks, make refactors, and more. Start building with Devin below:

Devin is generally available today! Just tag Devin to fix frontend bugs, create first-draft PRs for backlog tasks, make refactors, and more. Start building with Devin below:

Cognition

1,903,670 views • 1 year ago

Super excited about HydraFusion in GitHub Copilot, and what it shows about the shift from model selection to model orchestration. By bringing together multiple models to plan, build, critique, and complete coding tasks, it can deliver outcomes at up to 67% lower cost. It’s a great example of the value of a heterogeneous model ecosystem, and how we’re continuing to advance the cost-to-outcome frontier.

Super excited about HydraFusion in GitHub Copilot, and what it shows about the shift from model selection to model orchestration. By bringing together multiple models to plan, build, critique, and complete coding tasks, it can deliver outcomes at up to 67% lower cost. It’s a great example of the value of a heterogeneous model ecosystem, and how we’re continuing to advance the cost-to-outcome frontier.

Satya Nadella

327,340 views • 1 month ago

Set up your new GitHub Copilot agent for success! 🚀 This video will show you how to: ✅ Define clear, well-scoped issues. ⚙️ Utilize Copilot custom instructions effectively. 💻 Ensure a proper development environment setup. 🔗 Grant tool access via the Model Context Protocol (MCP). Watch and then head over to Docs for more tips! 👇

Set up your new GitHub Copilot agent for success! 🚀 This video will show you how to: ✅ Define clear, well-scoped issues. ⚙️ Utilize Copilot custom instructions effectively. 💻 Ensure a proper development environment setup. 🔗 Grant tool access via the Model Context Protocol (MCP). Watch and then head over to Docs for more tips! 👇

GitHub

69,843 views • 1 year ago

Introducing Dreaming: Across sessions, Devin builds a memory graph of how you like to work At night Devin self-improves its memory to remove stale records and discover latent information We are creating an OSS standard called Agent Memory Repo:

Introducing Dreaming: Across sessions, Devin builds a memory graph of how you like to work At night Devin self-improves its memory to remove stale records and discover latent information We are creating an OSS standard called Agent Memory Repo:

Cognition

614,293 views • 8 days ago

Introducing Devin 2.2 – the autonomous agent that can test with computer use, self-verify, and auto-fix its work. Try it for free! We’ve also overhauled Devin from the ground up: - 3x faster startup - fully redesigned interface - computer use + virtual desktop ...and hundreds more UX and functionality improvements.

Introducing Devin 2.2 – the autonomous agent that can test with computer use, self-verify, and auto-fix its work. Try it for free! We’ve also overhauled Devin from the ground up: - 3x faster startup - fully redesigned interface - computer use + virtual desktop ...and hundreds more UX and functionality improvements.

Cognition

541,026 views • 7 months ago

The terminal hasn’t changed much since the 1970s. What you do with it has. Introducing Devin for Terminal: everything we learned building Devin, now as a local agent, available right in your shell. And when your work outgrows your laptop, hand it off to the cloud.

The terminal hasn’t changed much since the 1970s. What you do with it has. Introducing Devin for Terminal: everything we learned building Devin, now as a local agent, available right in your shell. And when your work outgrows your laptop, hand it off to the cloud.

Cognition

394,379 views • 5 months ago

Cloud agents work while you eat, sleep, and take the subway. "Every morning I'll spin up a session on the way to work." Melius is a small and nimble startup, and Devin gives them the engineering capacity of a 100-person team. That's why the fastest moving startups use Devin.

Cloud agents work while you eat, sleep, and take the subway. "Every morning I'll spin up a session on the way to work." Melius is a small and nimble startup, and Devin gives them the engineering capacity of a 100-person team. That's why the fastest moving startups use Devin.

Cognition

122,114 views • 2 months ago

Gavin Baker says the future for the world's biggest companies is open models and private context: "I think the future is an ensemble of models. There's a Pareto curve. No one model is going to be the best at everything." "For the global 1,000 biggest companies, you're going to take whatever the best open-source model is. I think probably in the very near future, that's going to be an Nvidia model." "Everybody says 'Well, in a world where open-source wins, who funds the training?' Chip companies can fund the training." "It's trivial to do a $50 to $100 billion training run for Jensen, but I do think you're going to see American open-source led by Nvidia get really close to the frontier." "Sharing your own enterprise context, that's truly your IP, that's truly the value of your company, the context embedded in all of your data, sharing that with a frontier lab may be hazardous for your financial health." "I think what you'll see these companies do is they'll have their own model on their data, and it will work with one or two other frontier models, checking each other... I think that feels like a very likely future to me." Gavin Baker David George

Gavin Baker says the future for the world's biggest companies is open models and private context: "I think the future is an ensemble of models. There's a Pareto curve. No one model is going to be the best at everything." "For the global 1,000 biggest companies, you're going to take whatever the best open-source model is. I think probably in the very near future, that's going to be an Nvidia model." "Everybody says 'Well, in a world where open-source wins, who funds the training?' Chip companies can fund the training." "It's trivial to do a $50 to $100 billion training run for Jensen, but I do think you're going to see American open-source led by Nvidia get really close to the frontier." "Sharing your own enterprise context, that's truly your IP, that's truly the value of your company, the context embedded in all of your data, sharing that with a frontier lab may be hazardous for your financial health." "I think what you'll see these companies do is they'll have their own model on their data, and it will work with one or two other frontier models, checking each other... I think that feels like a very likely future to me." Gavin Baker David George

a16z

39,947 views • 1 month ago

underrated gems in Kimi-K3 release: > an early K3 wrote the majority of the kernels in the late development stages > it built a triton-class compiler from scratch, MiniTriton, that delivers performance on par with or better than Triton and torch.compile > then it designed a chip, by a model, for a model, in one 48-hour autonomous run the model is rewriting and optimizing every layer of the stack it runs on: kernels for its own training. A compiler for its own kernels. silicon for its own weights. we're watching software build its own hardware

underrated gems in Kimi-K3 release: > an early K3 wrote the majority of the kernels in the late development stages > it built a triton-class compiler from scratch, MiniTriton, that delivers performance on par with or better than Triton and torch.compile > then it designed a chip, by a model, for a model, in one 48-hour autonomous run the model is rewriting and optimizing every layer of the stack it runs on: kernels for its own training. A compiler for its own kernels. silicon for its own weights. we're watching software build its own hardware

Alex Ker 🔭

311,780 views • 2 months ago