1. Models.dev: Searchable database of AI model specifications and capabilities, Provider-by-provider pricing and model availability comparisons, Model pages with context limits, output limits, release dates, weights, and supported features, Provider pages listing available models, Laboratory pages grouping canonical models by author, JSON APIs for model, provider, and combined catalog data, Type-safe SDK with an offline snapshot, Provider and laboratory logo resources.
2. Bitrig: Generate native Swift apps from natural-language descriptions, Create full Xcode projects using the latest SDKs and Apple toolchain, AI-powered testing that taps, scrolls, types, detects issues, and fixes bugs, Built-in photorealistic 3D simulators for interactive app previews, Support for iOS, iPadOS, macOS, watchOS, tvOS, and visionOS, App Store and TestFlight archiving, uploading, signing, and publishing, App Store listing management and localized screenshot capture, GitHub integration for collaborative development workflows, CloudKit and Supabase backend support, Widgets, Live Activities, Controls, App Intents, Shortcuts, Siri, and Spotlight integration, In-app purchases, subscriptions, paywalls, StoreKit, and RevenueCat integration, Remote development and app distribution through iPhone and iPad.
3. Emdash: Parallel agent orchestration in isolated Git worktrees, Auto-detection of 25+ coding agent CLIs (Claude Code, Cursor, Codex, Gemini, etc.), Model Context Protocol (MCP) server integration, Built-in file editor and diff viewer, Issue integration with Linear, Jira, GitHub, GitLab, and Asana, Ephemeral infrastructure for cloud workspaces (Bring Your Own Infra via SSH).
4. Agen: Autonomous AI coding agents that work without constant prompting, Parallel agent execution — run multiple tasks at the same time, Full codebase awareness and context across your project, Supports real-world dev workflows: features, bug fixes, refactors.
5. Crew44: Multi-agent local workspace orchestrating teams of specialist agents working in parallel, Compounding skills and long-term project memory stored as plain files on disk, Model binding optimization allowing different AI models per role and provider hot-swapping, Companion mobile client access via an end-to-end encrypted tunnel, Local-first architecture ensuring no cloud accounts, no subscriptions, and no remote data tracking.