AI assistants and search decision
Aight
A competent developer can reproduce a useful client for local agents (sessions, multi-agent threads, voice, streaming) with several weeks' work, but matching the native polish, multi-device sync, App Store distribution, and built-in relay/service experience of the paid product is non-trivial.
Visit website↗$4.99/mo
$60/yr
Read off the official pricing page.
$100one-off80 h to build
$10/mo8 h/mo upkeep
On cash alone, building overtakes the subscription at 4 seats.
Open-source builds that already do this
Every project below is open source and already does this job today. Fork one, self-host it, or take the parts you need - the build prompt further down assumes an empty file, and this is the shortcut past that. Licences differ; check the one on each card before you ship. All Aight alternatives, with the arithmetic →
What a replacement has to do
- Connect to a running local AI agent, open or create session threads, exchange messages (with tool calls visible), orchestrate multiple agents in group threads, optional voice input/output and local TTS/STT.
What it still won’t have
- App Store distribution and associated polish (code signing, review, platform UX)
- Polished multi-device sync and account management
- Built-in encrypted Channels relay service (would need to implement or rely on third-party)
What remains hard
- Product polish and ongoing maintenance
First-year cost
Keep paying
Paying is—cheaper in year one.
On cash alone, building overtakes the subscription at 4 seats.
Money you would actually spend
Time you would spend
—
What you would spend
What we assumed
The verdict above measures whether you could build it. This one is only about money.
Runnable build prompt
Build a cross-platform client (Electron for macOS + a lightweight iOS SwiftUI companion or a progressive web app + iOS wrapper) that connects to local AI agent gateways (HTTP+WebSocket). In scope: implement agent discovery (mDNS/HTTP) and optional encrypted relay, session CRUD with local SQLite storage, searchable sessions, multi-agent group chats with routing and visible tool-call events, streaming responses via WebSocket, push-to-talk voice capture, local STT via whisper.cpp, local TTS playback (Soprano/PocketTTS) and a simple per-session token usage dashboard. Out of scope: training models, implementing an agent or model host, App Store release pipeline, enterprise multi-tenant server. Require robust error handling, connection retry/backoff, encrypted storage for credentials, and unit/integration tests for discovery, session storage, and message streaming.
How we checked
How the score was reached
- Partly verdict base52
- An open-source build was found+5
- 3 cited sources+3
- Price verified on pricing page+3
- 3/3 assessment runs agreed+4
- Evidence score67
The base comes from the verdict. Everything under it is a check that either happened or did not, and each one is a fact frozen in this record rather than a judgement made at render time - so the same evidence always produces the same number.
How scoring works →Cited sources · 3
Every page the run actually retrieved.
- official productAight — The App for Your AI Agents | Claude Code, Codex
- open sourcekhoj-ai/khoj
- open sourcearc53/DocsGPT
Integrity checks
What held up, and what did not.
