The NamiFusion Agent Canvas is an infinite node workspace where you chain models and processing steps into a single visual pipeline: text → image → edit → upscale → video, all connected. Describe a whole pipeline in one sentence and an in-canvas agent lays out the nodes and runs them. You can also drive the canvas from your CLI or coding agent (Claude Code, Codex, Gemini CLI) over MCP.
One visual pipeline
Chain generation, editing, upscaling and video on one canvas.
Natural-language agent
Describe a multi-step pipeline in one sentence; the agent builds it.
Drive it over MCP
Orchestrate the canvas from Claude Code, Codex or Gemini CLI.
How it works
- 1Start from a template or a blank canvas.
- 2Add and connect nodes — or ask the agent to build the pipeline.
- 3Run the whole chain and export the result.
NamiFusion 与主流 AI 工具对比
| NamiFusion | Midjourney / DALL·E / Firefly | |
|---|---|---|
| 内容过滤 | ✓无——如实渲染你的 prompt | 屏蔽或改写 prompt |
| NSFW 友好 | ✓是（仅受适用法律约束） | 禁止 |
| 水印 | ✓下载无水印 | 视情况 |
| 定价 | ✓按量付费 · 积分永久有效 | 订阅制 |
| 模型 | ✓一个账号，多家顶级模型 | 单一模型 |
| API 与 Agent | ✓统一 REST API + MCP | 有限或无 |
FAQ
Is this a ComfyUI alternative?+
It is a node-based visual pipeline with a natural-language agent and MCP orchestration, aimed at both creators and coding agents.
Can my coding agent drive it?+
Yes — the canvas is exposed over MCP, so Claude Code, Codex or Gemini CLI can orchestrate it.
Do I need to know node graphs?+
No — start from a template or describe what you want and let the agent lay out the nodes.