Coze

What it does

Coze is an AI application development platform from ByteDance. You design agents, workflows and complete AI applications using a visual canvas, and the platform runs them on large language model capabilities. Workflows take defined input parameters at a start node and return results through an end node. Outputs range from research reports and posters to picture books, or conversational responses for chat-based projects.

Builders drag and drop nodes onto a canvas, connect each node's output to the next step, and debug in real time to see data flow and execution order. Nodes cover LLMs, plugins, code, SQL, data add/query/update/delete, loops, batches, conditions and variable assignment, with types such as string, number, object, file and array. A plain workflow runs a task start to finish, while a chatflow handles multi-turn conversation logic for assistants or customer service. Agents can call plugins or workflows automatically.

It suits beginners and developers who want no-code or low-code ways to prototype and deploy LLM-based assistants and tools, including teams sharing a workspace. Workflow owners can edit, publish and delete their workflows; team members can view and use them. Pricing depends on the individual features used, and rates were not confirmed. Note that agents cannot be bound to chatflows containing chat nodes.

What it can do

• Drag-and-drop visual workflow canvas with real-time debugging
• Workflows for automated tasks and chatflows for conversational logic
• Nodes for LLM, plugin, code, SQL, loop, batch and condition steps
• Built-in data nodes to add, query, update and delete data
• Agents that call plugins or workflows to run business processes
• AI applications with business logic and a visual user interface

Example workflows

Generate an industry research report

Inputs: a topic defined as a start-node parameter. Steps: add LLM and plugin nodes, pass outputs between nodes, test on the canvas. Output: a report returned by the end node.

Build a customer service chatflow

Inputs: user chat messages. Steps: create a chatflow, add condition and data query nodes, bind it to an application. Output: conversational replies following the defined business logic.

Before you choose

• Pricing is per feature with separate billing models; check the Billing overview before relying on it
• Agents cannot be bound to chatflows that include chat nodes
• Workflow deletion by team members is limited to the team owner and administrators

Listed under

• Coding
• App builders
• AI agents
• AI app developers
• API
• App development
• Chatbot builder
• Chatbots

Get started

Start from the website or duplicate a template

1. Open the Coze website and create an account
2. Follow the quickstart chatbot or duplicate a template
3. Create a workflow or chatflow on the canvas, link nodes, and debug before publishing

Pricing is based on the individual features used, each with its own billing model and price. Specific prices, free tier and trial were not confirmed.

We couldn’t find published prices for Coze. See their pricing page ↗

Sources for this overview

Product pages, documentation, and pricing checked 29 Oct 2026. Each source supports part of the overview; availability and prices can change.

• Workflow overview - Document - Coze ↗docs.coze.com
• What is Coze - Document - Coze ↗docs.coze.com

Company website

coze.com: product, pricing and policy pages

GitHub

Public code and releases

YC directory mirror

Batch and company name

Web search

Confirms the official site

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

Last checked 29 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.