Coda was renamed Superhuman Docs in 2026, making Docs AI the current product surface corresponding to Coda AI. It puts generative AI inside a programmable document: AI can draft prose, but it can also create tables, formulas, buttons, views, dashboards, and workflows from a prompt. This is useful for teams willing to maintain shared data structures. It is unnecessary overhead for someone who only wants a general chatbot or a conventional text document.
Quick Verdict
Existing Coda customers should evaluate continuity under the Superhuman Docs name rather than treating it as a separate migration. New teams should prototype one real tracker and inspect the generated schema, formulas, permissions, and automation behavior after the demo effect wears off. Docs differentiates through structured building and more than 800 integrations. Notion AI emphasizes knowledge and writing, ClickUp Brain emphasizes project execution, and monday AI emphasizes standardized visual workflows.
Best For
Docs AI fits product operations, program offices, marketing operations, sales operations, and teams able to designate a few skilled makers who maintain templates and data models while others collaborate. It is less compelling for organizations unable to govern formulas, Packs, buttons, and source permissions, or those requiring behavior-preserving export to open formats. Procurement should not assume that historical Coda Maker Billing maps directly to the current Superhuman Suite member plans; existing contract entitlements need written confirmation.
Key Features
- Docs AI assistant: creates and summarizes documents, project plans, and tables, then answers questions over accessible content.
- Data and interface generation: builds trackers, dashboards, filters, formulas, buttons, and lightweight internal tools from prompts. A maker still owns the resulting schema.
- AI Views, Pages, and Databases: current plan documentation marks several of these as beta or limited trial, so critical records need stability and export testing.
- Connectors and two-way sync: brings systems such as Jira, GitHub, Figma, and Salesforce into Docs. Write-back Packs expand both convenience and blast radius.
- Docs MCP: permits clients such as ChatGPT, Claude, and Cursor to read or write Docs. OAuth scope, client retention, and write approvals must be reviewed together.
- Governance: document, page, folder, sharing, locking, and version-history features vary by plan; full SAML SSO, SCIM, audit logs, and organization controls are Enterprise capabilities.
Use Cases
- Generate a launch plan and responsibility tracker from a brief, then have the project owner repair dependencies and dates.
- Keep meeting notes, decisions, voting, and action items in one doc while AI summarizes only permitted content.
- Classify feedback row by row, sampling output before bulk execution to control both error and quota consumption.
- Build a management view over Jira or CRM data while restricting service-account scope and write-back actions.
- Connect an external AI client through MCP in a non-sensitive test workspace before granting production write access.
- Maintain a field dictionary and export schedule so the team can reconstruct essential records outside Docs.
Pricing
The current public purchase surface is the Superhuman Suite, not the old fixed Coda AI credit top-up table:
| Plan | Public price on 2026-07-21 | Docs boundary |
|---|---|---|
| Free | $0 | Collaborative docs and low automation quota; limited trials or beta access for Docs AI, databases, AI views, and MCP |
| Pro | $12/member/month annually; $15 monthly | Unlimited doc size, 30-day history, AI templates and insights, medium automation quota |
| Business | $33/member/month annually; $40 monthly | Unlimited automation and history, hourly connector refresh, plus other Suite products including Mail |
| Enterprise | Custom | SAML SSO, SCIM, audit logs, organization controls, custom roles and permissions |
The earlier 6,000/12,000 credits per Doc Maker and fixed top-up figures should not drive a new purchase. Existing Coda customers may retain contractual terms. Confirm member count, AI quotas, automation, connector refresh, Suite bundling, and renewal migration in writing.
Pros
- AI creates structured tables, formulas, views, and tools, not only prose.
- Documents, databases, collaboration, and automation share one surface.
- More than 800 integrations and MCP provide broad context and action paths.
- A free plan supports low-risk prototyping.
- Existing Coda workflows retain familiar concepts after the rename.
Cons
- Recent rebranding and packaging make old tutorials, domains, and credit numbers misleading.
- Formulas, Packs, buttons, and cross-doc relationships create strong behavioral lock-in.
- AI quotas, automation, and connector refresh vary by plan and need workload testing.
- MCP and two-way sync widen data and write access to external clients.
- Several newly emphasized data and AI surfaces remain beta or limited trial.
Alternatives
| Product | Better fit | Trade-off versus Docs AI |
|---|---|---|
| Notion AI | Wiki, writing, and knowledge Q&A | Polished content workflow; less programmable document behavior |
| ClickUp Brain | Tasks, projects, and agents | Deeper execution controls; less free-form app building in docs |
| monday AI | Departmental templates and visual process | Easier standardization; different formula and narrative flexibility |
| Airtable AI | Database-first apps and interfaces | Cleaner data model; weaker long-form collaborative narrative |
| Teable | Open-source spreadsheet database and self-hosting | More deployment choice; less document narrative and Suite integration |
FAQ
Was Coda AI renamed?
Yes. Coda became Superhuman Docs in 2026, and Docs AI is the current AI surface. Legacy coda.io pages point toward the new product.
Is pricing still based only on Doc Makers?
Historical Coda plans used Maker Billing, while the current public Suite page lists per-member pricing. Existing contracts may differ, so verify the actual account and quote.
Can Docs AI search an entire workspace?
Its useful scope should follow document, page, and folder access, while each Pack and MCP client adds separate authorization. Test with representative roles, not only administrators.
Can AI fully automate table processing?
It can generate, classify, and extract at scale, but makers still manage schema, exceptions, repeated runs, and quota. Put suggested output in a review column before consequential writes.
How can a team reduce lock-in?
Maintain a data dictionary, document important formulas, inventory connectors, export regularly, and avoid keeping the only approval evidence inside non-portable buttons or automations.
Is MCP secure by default?
No. MCP is a connection protocol. Security depends on OAuth scope, the external client, model data policy, token lifecycle, and which read/write tools are enabled.
Bottom Line
Superhuman Docs AI is valuable because AI works with document structure, data, and team collaboration instead of living in an isolated chat. That same depth creates permission, maintenance, quota, and migration responsibilities. Build one genuine but non-critical workflow, measure output quality and sync behavior, inspect exports, and price current member plans. It becomes a reliable work system only when the team is prepared to govern the model underneath the document.