Imitation AI

Release Notes

What's new in Imitation AI.
Every update, documented.

v1.5.3April 26, 2026

GPT-5.5 + DeepSeek V4 Flash

  • ·GPT-5.5 (OpenAI) — stronger reasoning, 1M+ context
  • ·DeepSeek V4 Flash — fast inference, 1M context
  • ·20 models from 10 providers
v1.5.2April 20, 2026

Kimi K2.6

  • ·Kimi K2.6 (Moonshot) — multimodal coding, Go support, agent swarm
  • ·20 models from 10 providers
v1.5.1April 16, 2026

Claude Opus 4.7

  • ·Claude Opus 4.7 — high-resolution vision, stronger agentic context
  • ·20 models from 10 providers
v1.5.0April 14, 2026

Light Mode, Mobile Navigation & Templates

  • ·Theme switcher — system, light, and dark mode for all public pages
  • ·Mobile navigation redesign — fullscreen menu with slide-down animation
  • ·Board Templates — Content, Research, and Code starter workflows
  • ·Minimap toggle in studio
  • ·7 blog articles published
  • ·Connect page — personal builder profile
  • ·WebP image migration for faster page loads
  • ·Mobile hero spacing aligned across all pages
  • ·19 models from 10 providers
v1.4.3April 7, 2026

GLM-5.1

  • ·GLM-5.1 (Zhipu) — strong coding model, structured output
  • ·19 models from 10 providers
v1.4.2April 2, 2026

Gemma 4 31B

  • ·Gemma 4 31B (Google) — open-source, 256K context, Apache 2.0
  • ·18 models from 10 providers
v1.4.1March 29, 2026

New Models & Security Hardening

  • ·xAI Grok 4.1 Fast + 4.20 Beta — new provider on the canvas
  • ·GLM-5 Turbo (Zhipu) — fast agent-optimized model
  • ·MiniMax M2.7 — upgraded to latest version
  • ·Gemini Flash + Pro — now working correctly
  • ·17 models from 10 providers
  • ·Clearer feedback when a model is unavailable or something needs attention
  • ·ImageNode now resizes vertically like other nodes
  • ·Toolbar works better on iPad portrait
  • ·Pricing page: correct tablet layout
  • ·Two new blog articles published
  • ·Performance improvements under the hood
v1.4.0March 19, 2026

Editable Output, Blog & Use Cases

  • ·Editable Output — edit AI output before it flows to the next node
  • ·Blog launched with first article: What is Visual Prompt Engineering?
  • ·Use Cases page — real workflows for content, research, and code
  • ·Welcome Modal — new users get a quick intro on empty boards
  • ·Code Node now uses parent context correctly in chains
  • ·Touch-friendly node handles for iPad and mobile
  • ·Better code readability in node outputs
  • ·SEO improvements — OG images, canonical URLs, Twitter Cards
v1.3.1March 18, 2026

Security & Stability

  • ·OAuth login now works on the first attempt — proper server-side code exchange
  • ·API error responses no longer leak internal details
  • ·Server-side model validation prevents unauthorized model usage
  • ·Stripe checkout redirect fixed for production environment
  • ·Cursor jump fix in all node types — edit text anywhere, not just at the end
  • ·Footer, favicons, and social sharing metadata improved
v1.3.0March 11, 2026

Launch Polish

  • ·Redesigned pricing — Think (Free), Prompt (€29), Evolve (€59)
  • ·Yearly billing option with 2 months free
  • ·New image carousel on the features page
  • ·Auto i/ai — let the Machine select the best model for your prompt
  • ·Installable as an app on your device (PWA)
  • ·Cleaner navigation — Releases now in the main menu
  • ·Under the hood: removed legacy code, unified branding, faster load
v1.2.1March 9, 2026

Smarter Nodes + New Models

  • ·Node output now scrolls — long responses stay contained instead of stretching the canvas
  • ·Input fields grow with your text — no more cramped prompts
  • ·Output renders Markdown — bold, headings, lists, and code blocks look great
  • ·Copy output with one click — just hit the copy icon
  • ·Scroll inside nodes without accidentally zooming the canvas
  • ·All node types are now resizable — drag to fit your workflow
  • ·Added GPT-5.4, GPT-5.3 Chat, and GPT-5.3 Codex to the model selector
v1.2.0March 8, 2026

Welcome to the Machine

  • ·Password reset — request a new password directly from the sign-in page
  • ·Email confirmation for new accounts
  • ·Profile settings — set your display name, company, and manage your account
  • ·Refreshed sign-in experience
  • ·Settings pages redesigned for clarity
v1.1.0March 1, 2026

Your Keys, Your Models

  • ·Bring your own API key — connect to OpenRouter and use any supported model
  • ·Usage dashboard — track your daily token usage at a glance
  • ·File uploads — attach TXT, CSV, PDF, DOCX, or XLSX directly to a prompt
  • ·Google sign-in
  • ·Collapsible sidebar for more canvas space
v1.0.0February 15, 2026

Initial Release

  • ·Visual prompt engineering on an infinite canvas
  • ·Four node types: Prompt, Note, Image, and Code
  • ·15 models from 9 providers — Claude, GPT, Gemini, Mistral, DeepSeek, and more
  • ·Connect nodes to chain your thinking — context flows automatically
  • ·Run Chain — execute your entire graph in one click
  • ·Boards save automatically