Imitation AI

Imitation AI is live. Build AI workflows visually. Connect prompts. Chain results. Think in graphs. Not a chat interface. A canvas. imitation.works

This week alone: GLM-5 Turbo (Zhipu) MiniMax M2.7 Grok 4.1 + 4.20 (xAI) Gemma 4 31B (Google) 5 models in one week. The canvas keeps up.

Gemma 4 just dropped — already on the canvas. Google's strongest open model. 256K context. Apache 2.0. 18 models. 10 providers. One workflow. imitation.works

GLM-5 Turbo already runs on our canvas. GLM-5V-Turbo is next. imitation.works/models

New: Templates. Start with a workflow, not a blank canvas. → Content Creation → Research & Analysis → Code Workflow Three starting points. Ready to run.

If prompting feels inconsistent… it's not you. You're working in a system that doesn't keep state. imitation.works/blog/what-make…

This is the shift: From asking → to constructing. That's what Think · Prompt · Evolve means.

Most AI tools are still chat. We didn't build chat. We built something else. New: "What Makes Imitation AI Different."

Model update. Grok 4.1 + 4.20 (xAI) GLM-5 Turbo (Zhipu) MiniMax M2.7 17 models. 10 providers. One canvas. imitation.works/models

"I don't trust anything other than myself to juggle this." That's why we don't pick your model. You do. 17 models. 10 providers. Your choice.

Visual workflows fix this. Each step becomes structure. Each output becomes context. Edit one node. Not the entire chain. imitation.works/blog/why-promp…

Chat is a conversation. Real work is a system. Content → draft → variations → translation Research → extract → compare → synthesize

A single prompt works. Prompting breaks the moment you need a system. New on the blog: imitation.works/blog/why-promp…

Content. Research. Code. Three workflows. One canvas. imitation.works/use-cases

Describe a feature. Generate code. Review it. Refactor. Three nodes. One flow.

Risk → Strategy → Decision. Three nodes. One workflow.