Loading…
AI automation & chat providers
One agent, every channel. Smart routing, chat providers, and AI in one command.
git clone https://github.com/iqqipro/nanoiqqi.gitConnect chat, AI, and tools through iqqi. Data flows in and out via one hub — the green node is iqqi at the center. No vendor lock-in, local-first optional, and one command to run.
Telegram, WhatsApp, Slack, Discord, Feishu, and more feed into one inbox. Notes (Notion, Obsidian), tasks (Things, GitHub), and docs stay in sync with your agent.
Send one command or question; iqqi routes it to the right channel, model, or tool. Smart Router picks the best model by cost and capability so you save without thinking about it.
iqqi at the center of your integrations. All beams point to one agent. Run self-hosted or hybrid; connect OpenAI, Anthropic, OpenRouter, local models, and 9+ chat channels from a single codebase.
One agent. Every channel, model, and tool. Connect what you use.
Telegram, Discord, WhatsApp, Slack, Feishu and more. One agent, all channels.
Anthropic, OpenAI, Google, OpenRouter, local models. Your keys, your choice.
Notes, tasks, wikis — Notion, Obsidian, Things, GitHub and more.
Spotify, Sonos, Shazam. Control playback and multi-room audio.
Philips Hue, Home Assistant, 8Sleep. Lights and IoT, voice-controlled.
Wrench, scripts, MCP. Extend with your own tools and skills.
9+
Chat channels
15+
AI Models
9+
Productivity tools
MIT
License
Pick the right model per request — save up to ~90% on token cost
In nanoiqqi, the Smart Router selects the best model per request using requirement detection, hard gates, an optional complexity judge, and multi-dimensional scoring. It routes by capability, cost, latency, and reliability, with automatic fallback across providers when one is unavailable or too slow.
What it does: requirement detection (e.g. tool use vs plain chat), hard gates (block certain models by task), optional complexity judge for harder prompts, and provider-diverse fallback so you stay within budget without losing quality.
Up to ~90% token cost savings with Smart Router
With vs without Smart Router (illustrative)
Lower latency for simple tasks when routing to the right model
Distribution across providers (illustrative)
Requests distributed across providers (illustrative)
Charts above use static illustrative data. For implementation details and configuration, see the Smart Router docs.
Ready to run iqqi app?
Clone the repo or run with npx. Your keys, your choice.