Capabilities

Home gives you a yearly contribution calendar of your tokens, prompts, or chats; hover over any day to see its count. Switch to Top Models to see which models used the most tokens over the time period you choose. Search, your account menu, and theme controls are always a click away.

Start a chat with a Local Host model or any connected API model, watch the reply stream in, and stop or retry a run whenever you need to. When the model supports it, you can change its reasoning effort right in the composer. Attach files, connect a File Room, Cohort, or connected app, or associate a coworker with the chat.

Each reply can include an activity log showing the searches, pages, and documents behind it, along with how long it took. You can edit a sent prompt or regenerate a response, and every chat gets a short title automatically, so it’s easy to find, rename, pin, archive, or move into a project later.

Projects keep related work together. Expand a project in the navigation panel to see its chats, or open its page to start a new project chat and browse earlier ones. Your Developer and Work activity stays tied to the selected project, and projects can be renamed, pinned, archived, or deleted at any time.

Invite coworkers by email, accept or reject requests, and manage your list of connections. When someone shares a chat with you, accept the invitation to open it. In a shared chat, both of you can read the conversation and prompt its model; the owner can invite others, and anyone invited can leave.

A Cohort is a team of agents you arrange on a visual canvas. Pick roles from the agent-role library, name each agent, give it its own Local Host or connected API model, and decide who reports to whom. You can edit, duplicate, or remove agents and pan, zoom, or expand the canvas as the team grows.

Run the whole Cohort or a single agent, follow the progress live, and stop it at any point. Results are saved in a run chat, and you can connect a Cohort to a chat to use it from the composer.

Create reusable agents with a name, persistent instructions, a provider connection, and a model. Give an agent a task and it runs through the provider you chose; recent runs and their status are listed so you can jump straight to the chat with the result.

File Rooms hold the reference material you want your models to work with. Upload documents, code, or text files, or write a new note, then browse each room’s files with their type, size, and total storage. Manage who is listed on a room, and connect it to a chat so the model has that context without re-uploading anything.

Add your API provider connections and see at a glance whether each one is healthy or needs attention. From there you can check a connection, re-enter a key, see how many models it offers, and pick one for chat. Connected models are then available everywhere in the app that runs provider models.

Tools is a catalog of apps you can connect, browsable by popularity or category and fully searchable. Read an app’s overview, connect your account through its sign-in flow, and see which accounts are connected. Select a connected app for a chat’s next run, and approve each action the model proposes before anything happens in that app.

Developer is a code workspace for each project, with a file browser, editor, and content search. Check Git status or a diff, use the sandbox terminal, and run quick Build, Tests, or Git status actions.

The Developer Assistant answers questions about your code and can take the open file as context. Every proposed change, tool call, and command waits for your approval, and you can see its output and errors afterward. File checkpoints, project memory, and prompt templates are built in.

Run an audit scan of an approved project workspace and review the results over time, including how many files were scanned and how many findings are open or high priority. Each finding explains its severity, category, file, and line, with a recommended fix. Open the affected file, dismiss the finding, or mark it resolved.

Work is where you create and edit project documents, spreadsheets, and presentations. Import DOCX, XLSX, or CSV files, keep everything searchable, and restore any saved revision from its history. When you’re done, export a document as DOCX or a spreadsheet as CSV or XLSX.

You can also ask a connected model about a file or a selection. It proposes edits, and you decide whether to apply or reject each one.

Settings lets you tailor Omnesly to the way you work. Manage your profile, two-factor authentication, and default project; choose your theme, language, typography, interface scale, and motion; and set your default model, response style, creativity, context limits, and workspace instructions.

You can also tune connection timeouts and web-access policy, notifications, accessibility options like high contrast and focus indicators, and privacy controls for exporting data or clearing the cache. Usage Status tracks tokens, requests, success rate, response time, reported cost, and provider usage over time.

Wherever you are, you can search for pages and settings, switch between Chat, Work, and Developer, and change how the navigation panel looks. Saved account sign-ins, the theme, and sign-out are always within reach.