1/ Today we're introducing Cua-S1-4B-0.2, the first multimodal decision model trained with RLOO on live computer-use tasks, using task-completion rewards. Text and multimodal adapters are available under Apache-2.0:

2/ At each step, the CUA-S1 model receives the screen state, the task goal, and a fixed set of candidate actions. It returns one action. The environment changes, and the next step starts from the new state.

3/ The training recipe has two stages. Supervised training teaches the decision format. Agentic RL then runs the model in live cua-bench-basic environments, where only a completed task earns the environment's reward.

4/ On the same held-out tasks, Cua-S1-4B-0.2 completes 17/18 text episodes and 13/18 multimodal. Zero-shot djev completes 16/18 and 12/18.

5/ On a frozen GUI-360 split of 168 multimodal tasks, Cua-S1-4B-0.2 reaches 92.9% versus 60.1% for untrained djev. Both use the same inputs and scoring, with no accessibility tree.

6/ The training code and benchmark results are now merged into Cua. Explore the code: Get the Apache-2.0 text and multimodal adapters:

What you all are doing is way over my pay grade. So I asked Grok why this is important. Awesome. Its answer: +++++ It matters because it is a concrete step toward computer-use agents that can be trained, inspected, and run outside the largest closed labs. Most computer-use systems still treat the desktop as a chat problem: a large model looks at a screenshot, writes a plan, and hopes the next click works. Cua-S1-4B-0.2 is built as a specialist instead. At each step it gets the screen, the goal, and a fixed list of candidate actions, then returns one action. The environment changes, and the next step starts from the new state. That is closer to how real UI work happens: many small, checkable decisions rather than one long generated essay. The training claim is the other reason people noticed it. Cua says this is the first multimodal decision model trained with RLOO on live computer-use tasks, using task-completion rewards. In other words, the model is not only imitating labeled clicks. After a supervised stage that teaches the decision format, it is rolled out in live cua-bench-basic environments and only scored when the task actually finishes. RLOO is a relatively simple online RL method (REINFORCE with a leave-one-out baseline), so the recipe is easier to reproduce than full PPO-style agent training. The numbers are narrow but not empty. On a frozen 168-task GUI-360 multimodal split, with no accessibility tree, Cua reports 92.9% versus 60.1% for the untrained djev baseline. On held-out live episodes it reports 17/18 text completions and 13/18 multimodal. Those are small, vendor-run evaluations, and Cua itself marks the checkpoint as early research. Pagination is still weak, and at least one live multimodal task is unsolved. So this is not evidence that a 4B adapter replaces Claude or OpenAI computer-use agents. It is evidence that a small, open decision layer can be trained against real GUI outcomes and beat its own untrained baseline by a large margin. That combination is useful in practice. The adapters sit on a frozen Qwen3.5-4B base, ship under Apache-2.0, and are small enough to run locally. Cua’s larger project already supplies the missing pieces around the model: a driver for real desktops, sandboxes, and benchmarks that can generate training trajectories. If the field is going to move from “frontier model remote-controls your screen” to “cheap, specialized policies that operate software,” this is one of the first public artifacts that looks like that stack. The honest limit is transfer. Strong results on Cua’s held-out splits do not automatically mean reliable performance on arbitrary apps, operating systems, languages, or layouts. The importance is the recipe, not a claim that desktop agents are solved: open weights, live-task rewards, one decision per step, and a training loop other people can inspect.

Love how fast cua team innovate computer use !

Hell yeah!! Congrats on this one Team Cua! 🍾 Fantastically done as always!

Will this be a part of Hermes Agent?

your mouse just became legacy hardware

Amazing work, looking forward to tet your model in (upcoming) Image Jev Bench. Maybe I should start a Computer Use Jev Bench as well

ain't gonna lie we really do need a computer use model

re: Cua There is one particularly interesting implication for FirstMate / computer-use agents: the really powerful architecture may not be Jev VS CUA-S1. It could be Jev ABOVE something like CUA-S1. Sol / Qwen / DeepSeek novel planning ↓ Jev task/policy decision ↓ CUA-S1 concrete UI decision ↓ Cua Driver deterministic execution ↓ verification That is starting to look less like “an LLM controlling a computer” and more like an actual inference hierarchy, where the expensive model is invoked only when the lower layers genuinely don't know what to do. @kunchenguid @CompleteSkeptic @typesafeai

can't wait to give it a run

a 4b model doing computer use means my laptop fan filed for hazard pay

🤯 this is crazy powerful, love cua!

training on task completion is the part i care about. in my own computer-use loop the model rarely clicked the wrong thing. it said done before the screen had actually changed. curious if completion rewards cut down on that early done

This is a great direction. Not every action needs a giant model thinking through it from scratch. Smaller, purpose built decision models for specific tasks make a ton of sense! 🫡

Love the specialized CU models and workflows coming out! Curious what the memory profile of this model is for running locally?

CUA agent focused on task completing is worth testing . Intresting !

Cua team is cooking!

always cooking 🔥

This just get's better and better

Stand by. Gpus spinning to add you to leaderboard.

Boom! Let’s goooo

@grok what is the accuracy rate of cua vs jev?

This is huge! Fastest computer use ever

im so hypeeee

letsgoo super excited to play more with cua - this inspired me to make a launch video with opus 5.5, best one shot i’ve ever seen

Every single day you all are shipping something new working at cua must be exhilarating

Browser and computer use about to be smooth

how does this work

How do I use Hermes agent to use CUA’s new model? @grok

Cua-S1-4B-0.2's focus on live tasks with task-completion rewards is intriguing

held-out computer-use tasks will reveal whether rloo generalizes beyond training workflows

The one-action loop is the interesting bit. Less agent theater, more measurable task completion

The no-accessibility-tree result is the spicy part. Would love to see a live-site eval where the button moves after the screenshot. That's when browser agents start sweating.

That GUI-360 jump made me look twice. 60.1% to 92.9% without an accessibility tree is a big result. I'm especially interested in what happens after a wrong click. Recovery is where computer agents tend to get messy.

One-episode gains on 18 tasks. GUI-360's 93% vs 60% is the story.