Forma
API Documentation

Design Intelligence API

Real-time UI terminology translation and behavioral intelligence for AI-powered design tools. Powered by Llama 3.1 8B on AMD MI300X.

LIVE v1.0.0 Llama 3.1 8B AMD MI300X Private Beta
QUICK START

Try the API

One curl command. Real response from production.

POST /v1/translate
● LIVE
curl -X POST https://forma-production-c800.up.railway.app/analyze \ -H "Content-Type: application/json" \ -d '{"text": "sticky top bar with frosted blurry popup"}'
Response (200 OK)
~1.5s latency
{ "phrases": [ { "phrase": "sticky top bar", "term": "Sticky Navbar", "definition": "A navigation bar that remains visible at the top.", "category": "Layout", "alternatives": [...] }, { "phrase": "frosted blurry popup", "term": "Glassmorphic Popover", "definition": "A floating panel with frosted glass aesthetic.", "category": "Overlay", "alternatives": [...] } ], "latency": 1547 }
ENDPOINTS

API Reference

Production endpoints and roadmap.

POST /v1/translate
● LIVE

Translate vague UI descriptions into canonical design terminology. Returns matched phrases, canonical terms, definitions, and ranked alternatives.

GET /v1/intelligence/{term}
COMING Q3 2026

Retrieve behavioral intelligence for a canonical term. Includes acceptance rates, regional preferences, top alternatives users selected instead, and trend data over time.

{ "term": "Off-Canvas Drawer", "global_acceptance_rate": 0.823, "detections_last_30d": 14782, "top_alternatives": [ { "term": "Bottom Sheet", "selection_rate": 0.12 }, { "term": "Sidebar Navigation", "selection_rate": 0.06 } ], "platform_distribution": { "v0.app": 0.45, "cursor.com": 0.20, "lovable.dev": 0.15 } }
GET /v1/insights/by-site
COMING Q3 2026

Aggregate insights about which canonical terms are most-described per AI builder platform. Useful for component library prioritization and design system gap analysis.

GET /v1/insights/trends
COMING Q4 2026

Time-series data on UI terminology trends. Track which components are gaining traction, which are declining, and emerging patterns across the AI builder ecosystem.

PRICING

Plans for Every Scale

From individual designers to AI builder companies.

Free
Individual designers
$0/month
  • 100 translations/month
  • Chrome extension access
  • 60 canonical components
  • Community support
Pro
Power users + freelancers
$49/month
  • Unlimited translations
  • Priority AMD inference
  • API access (10K req/mo)
  • Custom term definitions
  • Priority support
Team
Design teams + agencies
$499/month
  • Everything in Pro
  • API access (100K req/mo)
  • Behavioral insights API
  • Team analytics dashboard
  • SSO authentication
Enterprise
AI builders + Foundation labs
Contactus
  • Unlimited API access
  • Behavioral data licensing
  • Dedicated AMD MI300X
  • Custom canonical terms
  • SLA + dedicated support
Request API Access → View Dashboard
PLATFORM

Built for Scale

INFERENCE
AMD MI300X · vLLM 0.17.1
Sub-2-second inference on production GPU. Llama 3.1 8B serving real translations.
AUTHENTICATION
Bearer Token (Coming Q3)
API key authentication via Authorization header. OAuth 2.0 planned for Enterprise.
RATE LIMITS
Tier-based
Free: 100/mo · Pro: 10K/mo · Team: 100K/mo · Enterprise: Custom
DATA RESIDENCY
US-West (Default)
EU and APAC regions available for Enterprise customers.