Agent service
From CNY 99
per seat / month
- Build Agents and workflows around specific tasks
- Connect models, enterprise knowledge, and tools
- Require human approval for key actions
- Share and reuse projects with your team
Pricing
Starting prices for Tabro services
These are the monthly starting prices for each service. See each product's details for the full range; enterprise custom solutions are quoted separately.
Document service
From CNY 399
per month
- Read PDFs, images, and scans with OCR
- Extract fields that matter to your process
- Flag gaps and differences for review
- Keep results linked to source documents
Video service
From CNY 199
per month
- Organize stories and content by project
- Keep text, images, and video in one canvas
- Link visual assets to the relevant content
- Track review status and feedback
GEO service
Custom quote
- Observe how AI platforms describe your brand
- Examine cited sources and factual differences
- Manage approved claims and accountable actions
- Verify changes under the same run conditions
The four, side by side
Compare each service by task, capabilities, deliverables, and implementation scope.
Swipe horizontally to view every service
| Compare | Agent Platform | Document Management | Drama | GEO |
|---|---|---|---|---|
| Starting prices | ||||
| Monthly starting price | From CNY 99per seat / month | From CNY 399per month | From CNY 199per month | Custom quote |
| Typical work | ||||
| Best-fit scenario | Team development and reusable business Agents | Document extraction, pre-review, and research materials | Stories, short-video projects, and content review | Brand facts, AI answers, and cited sources |
| Workflow | Define task → connect knowledge and tools → run Agent or workflow → review key actions | Bring files → OCR → extract fields → flag items for review | Set theme → organize text and visuals → link assets → review content | Observe → explain → govern → act → verify |
| You provide | Tasks, knowledge materials, and tool interfaces | Source documents, required fields, and review rules | Themes, story settings, text, images, and video assets | Target questions, platform scope, approved claims, and sources |
| Capabilities and review | ||||
| Core capabilities | Agents, workflows, knowledge and tool connections | OCR, field extraction, and AI review prompts | Story settings, creative canvas, and linked media | Cross-platform observation, source analysis, and claim governance |
| Human checkpoint | A person approves key business changes or publishing | A responsible person makes the final call on flagged material | Reviewers confirm content and leave feedback | Approvers own claims; assigned owners handle and close issues |
| Deliverables | Agent and workflow setup, task results | Key fields, content summaries, and review prompts | Project content, linked assets, and review feedback | Observation evidence, actions, and same-method rechecks |
| Records | Run and human approval records | Processing results and confirmation records | Review status, changes, and feedback | Observation, action, and validation records |
| Integration and scope | ||||
| Systems and sources | Models, enterprise knowledge, APIs, Skills, and MCP | Document sources and downstream business systems | Existing content or asset systems | Observation platforms, enterprise sources, and governance workflows |
| Usage boundary | Public cloud or deployment in your own environment | Document access scope and required output | Asset sources, content access, and delivery format | Platforms, markets, languages, and run conditions |
| What changes the quote | Tasks, Agents and workflows, tool access, deployment | Document types, extracted fields, system integrations | Projects, content types, and collaboration needs | Platforms, markets, question library, observation frequency |
| Agree before signing | Permissions, approval points, and acceptance criteria | Document permissions, reviewer, and output requirements | Asset sources, member roles, review, and delivery | Run conditions, claim owner, and issue closure criteria |
Before we talk, think through a few things
What job you're trying to solve, whether you have a redacted sample, which systems need connecting, any deployment limits, and who signs off. Sort those out and the first call gets straight to substance.