Version 1.0 Roadmap

This page tracks progress toward the OpenModels v1.0 release. Each section represents a milestone with individual tasks showing current status.

Status: Released

Status Legend:

• Done — completed and shipped
• In Progress — actively being worked on
• Planned — scheduled, not yet started

Phase 1: Registry Foundation — Done

Establish the public registry as the canonical source of truth for models, providers, and mappings.

• Define JSON Schema for models, providers, and mappings
• Create validation script with schema, duplicate, and referential integrity checks
• Set up GitHub Actions CI for automated PR validation
• Populate initial model definitions (60+ models across 15+ families)
• Populate provider definitions (30+ providers with verified endpoints)
• Add provider-model mappings with pricing data (80+ mappings)
• Verify pricing against official provider documentation
• Support extended pricing dimensions (cache, image, audio, reasoning)

Status: Done

Phase 2: API — Done

Build the REST API for model discovery, comparison, and provider lookup.

• GET /api/models — list and search models with sort options
• GET /api/models/:id — model details
• GET /api/models/:id/providers — providers for a model
• GET /api/models/:id/compare — compare providers
• GET /api/models/popular — relevance-ranked popular models
• GET /api/providers — list providers
• GET /api/providers/:id — provider details
• GET /api/stats — registry statistics (counts + last sync)
• GET /api/search — unified full-text search
• GET /api/search/index — lightweight client-side search index
• GET /api/health — system health check
• Caching layer with TTL-based invalidation
• Pagination and filtering across all list endpoints
• Rate limiting and request throttling
• OpenAPI 3.1 spec generation with Swagger UI

Status: Done

Phase 3: Telemetry — Done

Real-time monitoring of provider health, latency, and availability.

• Health probe scheduler (5-minute intervals)
• Latency probe scheduler (15-minute intervals)
• Time-to-first-token measurement
• Total response time measurement
• GET /api/telemetry/health/:provider_id endpoint
• GET /api/telemetry/latency/:provider_id endpoint
• GET /api/telemetry/ranked/:model_id — ranked providers by performance
• Historical telemetry data retention (30 days)
• “Unknown” status for providers with no/stale health data
• Health probe accepts 401/403 as reachable (no false negatives)
• Alerting for provider outages

Status: Done

Phase 4: Web Interface — Done

Web application for browsing, searching, and comparing models at openmodels.run .

• Model search and browse page with sort options
• Model detail page with provider list and error handling
• Provider comparison view (pricing, latency, uptime)
• Command Palette (Cmd+K / Ctrl+K) for instant global search
• Popular Models section with relevance scoring
• Provider logo system with SVG assets and letter fallback
• Provider Marquee (auto-scrolling logo strip)
• Ecosystem Graph (interactive node visualization)
• Recently Added section with relative timestamps
• Category-based navigation (capability, modality, license)
• Dark/light mode theming
• Responsive layout for mobile devices
• SEO optimization (meta tags, sitemap, structured data)

Status: Done

Phase 5: Documentation — Done

Comprehensive documentation site for users and contributors.

• Docs site setup with Nextra
• Architecture documentation (overview, data flow, schemas)
• Contributing guides (adding models, adding providers)
• Roadmap page
• Getting Started section (quickstart, concepts, installation)
• API Reference section with endpoint documentation
• SEO and sitemap configuration
• Deployment to production

Status: Done

Phase 6: Production Hardening — Done

Final polish and production readiness for v1.0.

• API rate limiting (100 req/min per IP)
• OpenAPI 3.1 spec auto-generation with Swagger UI
• Provider outage alerting (webhook + email)
• SEO optimization (structured data, meta tags, sitemap)
• Lighthouse scores > 90 (Performance, Accessibility, SEO)
• Security audit (no critical or high findings)
• Load testing (100 req/s sustained)
• Production deployment with zero-downtime updates

Status: Done

Release Criteria — Met ✓

• All Phase 1–6 tasks marked as done
• API uptime > 99.5% over 30-day window
• Documentation covers all public endpoints
• Lighthouse scores > 90 (Performance, Accessibility, SEO)
• Security audit passed (no critical or high findings)
• Load testing confirms API handles 100 req/s sustained