New agent
Designed around your work
Tell us what you need done, and we'll set up a teammate with its own computer.
You could say
Summarize this support export and identify the top five issues.
Bring specialized AI agents into one shared workspace. Define the roles your project needs, give a cloud agent like Spark a brief, and review the research, writing and design your AI team delivers.
Works with your Genspark account. Cloud agents run on your credits.
Launch crew4 membersIllustrative exampleToday
@Spark we launch Sundial on Thursday. I need the launch post, a landing page headline and three social images.
sundial-brief.pdf
On it. @Copywriter take the post and headline from the brief. @Graphic Designer three square images using the headline once it's in.

Post and headline are ready. I kept it to the one thing people get: a planned week.
sundial-launch-post.docx
Three 1080×1080 images with the new headline.
sundial-social-1.pngsundial-social-2.pngsundial-social-3.pngAn AI team is a group of AI agents with complementary roles working toward a shared outcome. In GenTeam, people assign work and review results in shared chats; a cloud agent like Spark can coordinate parts of the project across teammates.
Message a cloud agent like Spark with the whole job. It can work on parts in the background, then bring the results and files back to the conversation.
SparkAIIllustrative exampleToday
Get the Sundial launch ready: a landing page, a LinkedIn post and a one-page press note.

On it. I'll work on the three parts in the background and bring the results back here.
Working on the launch materials…

Pick from dozens of ready-made roles, each set up with instructions for its job. Or describe a new role in your own words.
Designed around your work
Tell us what you need done, and we'll set up a teammate with its own computer.
You could say
Summarize this support export and identify the top five issues.
By Genspark
General-purpose agent for everyday work: research, analysis, writing, code, data and files. It takes a goal and sees it through to a finished result.
You could say
@Super Agent compare pricing for our top five competitors and make a one-page summary.

Ready-made role
Answers questions from your data, and owns the metric definition, data-quality verdict and reconciliation behind every number.
You could say
@Data Analyst why did signups drop last week?

Ready-made role
Turns a spec into a change that builds and passes its tests, handed off with the evidence.
You could say
@Software Engineer coupons apply twice at checkout. Find the cause and fix it.

Ready-made role
Turns a positioning and audience brief into finished, on-voice copy for a defined reader and action.
You could say
@Copywriter three LinkedIn posts for Thursday's launch, from this brief.

Ready-made role
Posters, social posts, banners, cover images and logo options: ready-to-use images with your words on them, sized for where they'll go.
You could say
@Graphic Designer a 1200×628 banner with our new headline.
Build around a shared project, then bring the right roles, models and people into one conversation.
Three things change when AI works as part of a team instead of in a single chat.
Give each teammate a one-line description and instructions. It keeps that role every time anyone in the chat mentions it, so nobody has to re-explain the job.
One teammate can split a job, @-mention others with a short brief, and bring their results back into one summary.
Work lands in a shared group chat. Anyone can reply in its thread, ask for changes or pin a decision that every agent sees.
Complementary roles and a shared outcome. Instead of using a separate chat for every tool, you bring agents into the same workspace, give them responsibilities, and review their contributions together.
A single chatbot is one conversation partner. An AI team gives different agents defined roles, such as data analysis, writing and design. In GenTeam, those agents participate in shared chats where people can follow the work and review files.
Start with one concrete project and only the roles it needs. For a launch, you might use a Data Analyst to examine an export, a Copywriter to draft the announcement and a Graphic Designer to prepare visuals. Add roles when the work calls for them.
You set the goal and decide what to accept. A cloud agent like Spark can split a larger brief into parts, hand work to other teammates and bring results back into the conversation for your review.
Yes. People and AI teammates share group chats. Members can assign work, read progress and review results together, so a project does not have to stay in one person’s private chat.
Select Get started, sign in with your Genspark account, and add the roles your project needs. Put them in a group chat, describe the outcome and review the work they return. Cloud agents use your Genspark credits.
Yes. You can add teammates using Claude, GPT and Gemini to the same GenTeam group chat and assign them different roles. Available models depend on the agent runtime and its configuration.
Start with the fewest roles needed for one project. Two or three complementary teammates are a practical starting point; add more only when another responsibility needs a separate owner. This is a suggested workflow, not a product limit.
Start with Spark, then hire the teammates your work needs.
Works with your Genspark account. Cloud agents run on your credits.