AgentUI
Also known as: AgentUI, Agent UI, agentui.ai, AgentUI CLI
Managed AI app builder for operations teams that generates and hosts internal tools with a human support team behind them, audit logs, role based access, a SOC 2 badge and a JSON CLI for outside coding agents.
AgentUI is a managed AI app builder for operations teams that pairs its AI with a human support team. A builder describes the operation, such as inventory, approvals, reporting or multi location control, and the AI generates the screens, workflows, logic and database. Builders can also start from a gallery of 37 verified blueprints. The app goes live on AgentUI's hosting right away, and AgentUI's staff review the tool, connect the customer's systems and help as requirements change.
Agents read and write SQL databases, Salesforce, HubSpot, Shopify, Gmail and other services through an integration catalog with OAuth2 and API key authentication, and backend functions become tools an agent can call. Workflows start from incoming emails, webhooks, CRM events and Zapier triggers.
The platform's governance features are aimed at teams leaving spreadsheets. They include global audit logs of every edit, role based users, a compliance officer role, an internal portal that lists every tool, shared databases and shared secrets. AgentUI shows a SOC 2 badge, states GDPR compliance and runs a trust center, and Enterprise adds private deployment, SSO and security dashboards.
For developers, an AgentUI CLI with a published command reference returns JSON to outside coding agents so they can query an app's database, invoke a function and ship without relaying output by hand. AgentUI says there is no SDK and publishes no API reference. AI credits are spent only while building, and live apps run without consuming them. Plans start at $9 a month after a 14 day trial, with no free plan. AgentUI does not name the model behind its builder AI, and it names no testing or approval controls for the AI's changes.
Vendor details
Canonical URL
https://www.agentui.ai
Category
Coding agent
Subcategory
Managed AI app builder for operations
Funding status
Private. Funding is not stated on AgentUI's own pages.
Company status
independent
Use cases & customers
Primary use cases
Target customers
Deployment options
Integrations
Generated apps send email, use WhatsApp, call any external API and read and write existing SQL databases, with Shopify, Google Sheets, Odoo and Salesforce pages on the site. An AgentUI CLI returns JSON so outside coding agents such as Claude Code or Cursor can query an app's database, invoke a function and ship.
In practice
A distribution company describes its stock process and gets an inventory app with low stock alerts and movement history, then AgentUI's team connects it to the existing SQL database.
An operations manager replaces a spreadsheet of purchase approvals with an approvals app whose every edit lands in the global audit log.
A developer points Claude Code at the AgentUI CLI to query an app's database and ship a change.
Sources & related URLs
Related / legacy domains
Agentic Index coverage score
7.0 / 14 capabilities · 50%
| Integrations & Tool Calling | Full |
|---|---|
|
Agents read and write SQL databases, Salesforce, HubSpot, Shopify, Gmail and other services through an integration catalog with OAuth2 and API key authentication, and backend functions become tools an agent can call. Sourceagentui.ai/en/ai-agent-toolsread 2026-10-25 |
|
| Workflow Orchestration | Partial |
|
Inside one builder, the AI produces screens, workflows, logic and database for an app in multiple steps once the customer describes the operation, and apps can include backend functions. AgentUI names no multi agent execution or orchestration of the AI within workflows. SourceAgentUI, agentui.airead 2026-10-06 |
|
| Knowledge Grounding & RAG | Not documented |
|
AgentUI names no retrieval or grounding layer for the AI. Its help center has a collection on using skills, but it does not say what skills contain. Sourcehelp.agentui.airead 2026-10-02 |
|
| Human Oversight & Guardrails | Partial |
|
AgentUI's own team reviews generated tools, connects systems and supports the customer as the operation changes, and Team plans add a second environment. AgentUI names no step where the customer approves the AI's changes before they commit. SourceAgentUI, agentui.ai and help.agentui.airead 2026-10-06 |
|
| Security, Identity & Governance | Full |
|
AgentUI shows a SOC 2 certified badge, states GDPR compliance and bank level encryption, and runs a trust center. Users carry specific roles, including a compliance officer role, with shared secrets and audit logs. Admins create, edit and delete users with role assignments, and Enterprise adds SSO and security dashboards. The SOC 2 report itself is not public. SourceAgentUI, agentui.ai, /en/pricing and help.agentui.airead 2026-10-06 |
|
| Observability & Auditability | Partial |
|
Global audit logs record who changed what and when, with 7 day retention on the 25 and 50 member Team tiers and unlimited on Enterprise, plus custom audit logs inside apps. AgentUI describes no record of the builder AI's own steps. Sourceagentui.ai/en/pricingread 2026-10-02 |
|
| Memory & State Persistence | Partial |
|
The AI keeps conversation state within a chat, and AgentUI advises starting a new chat for each new task. No memory layer that lasts beyond a chat is named. SourceAgentUI, help.agentui.ai best practices for working with the AIread 2026-10-06 |
|
| Deployment & Data Residency | Partial |
|
Apps run on AgentUI's built in hosting and can sit on the customer's own domain. Enterprise adds private deployment alongside SSO, security dashboards and a dedicated support engineer. AgentUI does not say where or how private deployment runs, and it names no region choice or self hosted install. SourceAgentUI, agentui.ai/en/pricing and help.agentui.airead 2026-10-06 |
|
| Prebuilt Agents, Templates & Packs | Full |
|
AgentUI publishes a gallery of 37 verified blueprints that customers start an app from, and the CLI lists and creates apps from them, alongside use case pages for dashboards, inventory, approvals and reporting apps. Sourceagentui.ai/en/blueprintsread 2026-10-25 |
|
| Triggers & Channel Coverage | Full |
|
Trigger based automation starts workflows from incoming emails, webhooks, CRM events and Zapier triggers, so work starts without a person asking. Builders also reach the AI through chat, and outside coding agents reach it through the CLI. Sourceagentui.ai/en/skillsread 2026-10-25 |
|
| Model Flexibility & Routing | Not documented |
|
AI credits are priced from $0.40 a credit on Keep it running down to $0.25 on Builder and Team. AgentUI does not name the models behind its builder AI and names no model selection or bring your own key option. The Gemini, OpenAI, Claude and Bedrock integrations serve the apps it builds, not the builder. Sourceagentui.ai/en/pricing and /en/ai-agent-toolsread 2026-10-25 |
|
| APIs, SDKs & MCP Extensibility | Partial |
|
The AgentUI CLI, now with a published command reference, returns JSON for every command so an outside agent such as Claude Code or Cursor can query an app's database, invoke a function and ship without anyone relaying output. AgentUI says there is no SDK to adopt, and no API reference is published. The help center lists an API and Database section behind sign in. Sourceagentui.ai/agentui-cli.mdread 2026-10-25 |
|
| Testing, Debugging & Optimization | Not documented |
|
For the AI itself, AgentUI names no testing, evaluation or self check capability. AgentUI's human team reviews tools and makes sure they work, as part of its service. Operations teams report 67% better process efficiency, builds 2.3 times faster than custom ones and 44% less budget overage. SourceAgentUI, agentui.airead 2026-10-06 |
|
| Browser & Computer Use | Not documented |
|
The AI builds apps that reach other systems through email, WhatsApp, external APIs and SQL databases. AgentUI names no browser, desktop or remote computer control by the AI. SourceAgentUI, agentui.airead 2026-10-06 |
|
The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded
Pricing
Keep it running $9/mo · Solo $22.50/mo yearly ($25 monthly) · 14 day free trial
Flat plan by team size with monthly AI credits, and per credit rates by plan
Included quota
Keep it running: 3 apps, 3 members, 1,000 records, no included credits. Solo: 90 credits, 5 projects, 1 member. Builder: 200 credits, 10 projects, 2 members. Team: 400 credits for 10 members, 1,000 for 25, 2,000 for 50, unlimited apps.
What is public
All plan prices, credit allowances, per credit rates, member caps, the 14 day trial and a feature comparison are published. Enterprise is custom.
Billing mechanics
Monthly or yearly (10 percent off) plan fees sized by internal members and projects, with an included monthly credit allowance; live apps do not consume credits.
Cost watchouts
Credits are spent whenever the AI generates screens, logic or workflows, and the site estimates 90 credits at roughly 5 to 9 app screens. Audit logs are absent on Team for 10 members, kept 7 days on Team for 25 and 50 and unlimited on Enterprise, and internal users are capped by tier.
Variable cost rationale
Plan fees are fixed and published, and credits are metered at published per credit rates, but consumption per build varies with app complexity.
Additional watchouts
Credits are spent only during building. Inference, not stated by the vendor: the cost profile front loads into the first months of a project.
Overage / add-ons
Per credit rates are listed per plan ($0.40 Keep it running, $0.28 Solo, $0.25 Builder and Team).
Sales call required
Mixed (some tiers require a call)
Free / trial
14 day free trial on every workspace. After it, apps pause until a plan is chosen and are kept for 60 days. No free plan.
Lowest paid plan
Keep it running at $9 a month: 3 apps, 3 members, 1,000 records and no included credits ($0.40 per credit)
Commercial notes
Self serve through Team; Enterprise through sales with a dedicated support engineer. White glove onboarding is included.
Key ambiguities
Whether extra credits on paid plans are bought at the listed per credit rates is implied but not stated outright.
Missing data
Enterprise pricing is not published.
Related vendors
- Cognition — Maker of Devin, an autonomous AI software engineer
- 10Web — Agentic website platform whose AI agents build, host and manage…
- Aider — Open source, model agnostic terminal coding agent that edits your…
- Amp — Amp is a coding agent from Amp Frontier Corporation that runs each…
- Anthropic Claude Code — Anthropic's agentic coding system across terminal, desktop, IDE, web…
- Anything — Anything is an AI app builder whose agent turns a plain language…
Alternatives to AgentUI
The closest documented capability profiles to AgentUI among coding agents tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.
- Google AI Studio7.5 / 14Adds documented Knowledge Grounding & RAG and Testing, Debugging & Optimization
- Locofy7.5 / 14Adds documented Knowledge Grounding & RAG and Testing, Debugging & Optimization
- Rocket9.0 / 14Adds documented Knowledge Grounding & RAG and Testing, Debugging & Optimization, among others
- Graphite8.5 / 14Adds documented Knowledge Grounding & RAG and Model Flexibility & Routing, among others
- StackBlitz Bolt9.5 / 14Adds documented Knowledge Grounding & RAG and Model Flexibility & Routing, among others
- Blink10.0 / 14Adds documented Knowledge Grounding & RAG and Model Flexibility & Routing, among others
Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded