Addressable AI
Deploy nodes instantly
On-demand cognitive intelligence with any LLM.
Robust, configurable, and standards-based AI.
What is an Interlocute node?
Interlocute nodes are persistent, addressable units of intelligence. They handles inputs, manage memory, and execute logic instantly or asynchronously.
Same Models. Better UX.
Subscription chat services use the same LLM APIs you could call yourself. But because of subscription pricing, they throttle you whenever. Interlocute is for those that want to unleash the power of the models without worrying about the infrastructure. Interlocute uses the same models you know with bigger context, full observability, and a rich workspace you'd otherwise need a dozen libraries to build.
Full Context Windows
No silent trimming. You choose how much context your node receives.
Observable Processing
Token counts, latency, capability traces — see everything.
Configurable Everything
Model, persona, constitution, capabilities — all live, no redeploy.
Real Cost Accounting
Per-request metering. Per-thread cost attribution. No opaque tiers.
Rich Workspace
Threads, tagging, prompt saving, documents, artifacts — all built in.
No Assembly Required
One app replaces vector DBs, prompt tools, and billing libs.
How it works
Deploy your intelligence in three simple steps. No infrastructure management required.
Create
Initialize your node environment instantly using our CLI or web dashboard.
Configure
Choose a model, pick features, and configure settings.
Call
Chat through many built-in surfaces, use the Api, or set recurring tasks.
Key Capabilities
Engineered for Production
Runtime API
Low latency addressable node interactions with stateful memory.
Streaming
Native SSE support for real-time token delivery to front-end clients.
Scheduling
Built-in cron engine for proactive intelligence tasks.
Controls
Granular IAM policies to restrict node access by token or origin.
Out-of-the-box agents
Preconfigured AI agents you can deploy in minutes. Each agent is a proven bundle of capabilities — governed, metered, and production-ready.
Interactive Assistant
Conversational AI
A conversational AI assistant you can deploy in minutes. Threaded conversations, persistent memory, and real-time streaming — ready to chat.
- Set up in minutes — no infrastructure to provision
- Persistent conversation threads with long-term memory
- Real-time streaming responses (SSE)
Automation Worker
Scheduled automation
An AI agent that runs on schedule — daily digests, data processing, and monitoring without human supervision.
- Built-in cron-style scheduling — no external scheduler needed
- Trigger-driven execution from webhooks and events
- Background summarisation and data processing
API / Contract Node
Structured API endpoint
A strict AI endpoint for service-to-service calls. JSON-in, JSON-out — deterministic, governed, and production-ready.
- Strict JSON-in / JSON-out contract enforcement
- Minimal conversational overhead — no greetings or commentary
- Designed for service-to-service and webhook integrations
Media Processor
Media processing
An AI agent built for media — transcribe audio, analyse images and video, parse documents. Upload, process, and get structured results.
- Audio transcription (speech-to-text) out of the box
- Image upload with OCR and vision analysis
- Video ingestion and frame extraction
Built-in patterns for agentic builders
Infra-less intelligence primitives
Long-term Memory
Persistant user context across threads.
RAG (Knowledge Retrieval)
Plug in your docs; we handle the vector search and injection.
Tool Use
Pre-configured function calling for web search and data processing.
State Management
Manage complex multi-turn conversation states automatically.
Observability
Built-in logging and tracing for every node call.
Budgeting
Built-in protection to keep your costs and upstream usage safe.
Prompt Templates
Use our prompts or bring your own, all without redeploying code.
Streaming SSE
Real-time, character-by-character responses for a snappy UI.
Deep Reasoning
Multi-step plan, draft, critique, and revise — with inspectable trace artifacts.
Video Intelligence
Transcripts, scene analysis, entities, and AI summaries from video.
Document Processing
OCR, tables, form fields, and figures from PDFs and scanned docs.
Image Intelligence
Structural fingerprinting, semantic understanding, and forensic verification.
Talk to your agent anywhere
Connect your agent to the channels your users already use — web, widgets, messaging apps, and more.
Hosted Web
AvailableGive your AI agent its own web address. A full hosted chat experience — themed, branded, and ready to share.
Chat Widget
AvailableDrop an AI chat widget onto any website. One script tag, instant conversations, and full domain control.
Telegram Bot
Coming soonConnect your AI agent to Telegram. Users message your bot, your agent responds — conversational AI in the app they already use.
Let users talk to your AI agent on WhatsApp. The world's most popular messaging app, powered by your Interlocute node.
Configure via Conversation
Stop fighting with complex JSON schemas. Evolve your node's behavior by simply describing what you need.
Applying 'technical_writer' persona...
Configuring RAG tools for PDF parsing...
Built for these use cases
Real-world scenarios where Interlocute nodes excel. Each use case ships out of the box — no infrastructure to build, no pipelines to manage.
Embed Slim Chat via IFrame
Add a live AI chat to any website with a simple iframe embed. No backend code, no infrastructure — just paste and go.
Embed Chat Widget (JavaScript)
Drop a floating AI assistant onto any page with a single script tag. Customizable positioning, theming, and streaming responses out of the box.
Call Node as a Function (API)
Call an AI node from your server-side code to power product features, automate workflows, or process data with a simple REST API.
HTTP API Reference
Full REST API for chat, threads, and streaming. Build custom integrations with standard HTTP endpoints and predictable contracts.
Usage-based pricing. Start small.
Pay only for the tokens your nodes consume. No monthly platform fees.
"By enabling Intelligence to be Addressable, we believe a Commons will emerge. Like the World Wide Web before it, standards allow people and agents alike to flourish."
Built for builders
Our documentation is written by developers, for developers. Get up and running in less time than it takes to brew a coffee.
Authentication
Include your API Key in the Authorization header: