The Flutter renderer
for AI output.
Built for production Flutter AI interfaces. Render streaming assistant replies, Markdown, LaTeX, code, tables, citations, and custom inline UI with one widget.
AI output lab
See real AI output render
Compare incoming model text with the native Flutter experience your users receive.
## Release brief A streaming renderer should feel like a considered part of the interface, not text arriving in chunks. It keeps the readable rhythm intact as the answer becomes useful.
Built for production AI output
The rendering details that make assistant output feel native to your product.
AI-ready rendering
Markdown, LaTeX, code, tables, citations, RTL, and selectable text render together without preprocessing the model response.
Streaming that stays fast
Reveal replies with intentional pacing while only the unfinished tail rebuilds, then finish cleanly when generation stops.
Fits your design system
Twelve component style classes, app-wide theme support, builders, and callbacks give appearance and structure separate controls.
Extensible by default
Autolinks, inline patterns, and scoped custom components add app-specific UI without claiming the wrong text.
New in v1.2.0
A stronger rendering layer
v1.2.0 focuses on streaming behavior, design-system control, and safer extension points.
Streaming reveal
Split-document caching keeps settled content stable while the live tail rebuilds.
Component style sheet
Style Markdown parts per widget or app-wide without replacing their structure.
Inline patterns and scopes
Add mentions, channels, and custom components with explicit nesting rules.
Better edge-case handling
Autolinks, RTL ordering, text scaling, reduced motion, and runtime theme changes.
Give AI output the same care as the rest of your app.
Add one widget, keep control of the details, and ship a response surface that feels native on every Flutter platform.