Oh!Mate is a desktop AI workspace that lets users hand off office tasks—document editing, spreadsheet analysis, slide creation, and file organization—to an agent that plans the work, requests permission at key steps, and returns editable files. It targets professionals who deal in Word, Excel, PPT, and PDF files daily and want an agent to carry a task from source material to a finished deliverable without losing track of what happened along the way.
The core unit in Oh!Mate is a Work, which bundles the authorized source files, the conversation, the agent's plan, permission records, and the resulting deliverables. Users define a source scope by selecting files or folders, describe a goal, and let the agent proceed. Because everything tied to a project lives in one Work, long-running jobs do not dissolve into scattered chat threads. Multiple Work tabs let users run parallel projects, and a local working directory keeps outputs on the machine.
Rather than acting as a single prompt-and-response box, Oh!Mate's agent breaks a goal into steps, lists the tasks involved, and explains the actions it intends to take. When a step requires reading, writing, or external access, the agent asks for confirmation. An activity stream records plans, file operations, commands, permission requests, and retries, giving users a traceable account of what the agent did and why. This design suits work where an unexplained file change would be a problem.
Oh!Mate handles the formats most office work depends on. For Word documents, it can read, rewrite, annotate, and check formatting. For Excel, it cleans and reconciles data, preserves formulas, and builds reports. For PowerPoint, it develops outlines, creates slides, and checks layouts. PDFs and images can be extracted, compared, and reorganized into new deliverables. The emphasis on original formatting matters for contracts, proposals, and reports that carry house styles or client templates.
The product's scenarios cluster around a few recurring jobs. Contracts, reports, and forms get read for key points and risks, with revisions saved to a separate directory so originals stay intact. Operational and financial data gets cleaned, checked for anomalies, and turned into analysis summaries with validation results attached. Proposals and presentations start from a shared outline before the agent drafts slides. Teams can keep project context in a Work so recurring updates, FAQs, and weekly reports draw on the same source material. Marketers and creators can adapt one source into platform-specific copy, scripts, and publishing checklists—useful for anyone who needs an AI copywriter's output shaped to different channels.
The product frames its value around three ideas: control, visibility, and traceability. Users see what the agent reads and changes, approve sensitive actions, and review an audit trail afterward. Work-log Skills extend this by capturing completed tasks, key decisions, blockers, and next steps, which helps when a project spans days or changes hands. The platform is described as built on open standards, with users able to choose their providers and workflows.
Oh!Mate uses a Credit system that resets every 7 days across all tiers. The Free plan provides 40 Credits for lightweight, occasional tasks. Plus at $15.00 per month offers 500 Credits for everyday document and content work, Pro at $30.00 per month provides 1,100 Credits for complex, multi-step tasks, and Max at $90.00 per month supplies 3,500 Credits for intensive delivery and parallel projects. Every tier includes the full set of office capabilities. Credits are consumed according to actual model usage, so the token mix of a given task affects consumption more than the number of tasks. Upgrades take effect shortly after payment and bill for a full period without proration; downgrades and direct moves to Free are not currently supported.
Oh!Mate is a reasonable fit for people whose work revolves around Office documents and who want an agent to move tasks forward while keeping a visible record of each step. The permission model and activity log address the trust gap that keeps some teams from delegating file work to AI. The tradeoffs are the desktop-only setup and a Credit system that rewards users who understand how model usage translates into consumption.