Platform overview

SAGEA has five products. Pick the one that matches what you want to do.

Helios: identity verification infrastructure. Liveness, document checks, and KYC via hosted APIs.
Studio: model APIs for chat, voice, and documents. Keys, Playground, and SDKs to ship applications.
MYLO: voice agent solution. Real-time conversational agents with Sonus voices and SAGE reasoning.
RUNE: local-first coding agent. Natural-language codebase help with models running on your machine.
Foundry: on-prem solution. Run SAGEA models inside your own infrastructure. No hosted APIs.

For model help, see Choosing a Model. For costs, see the comparison table there.

Helios

Helios is SAGEA's identity verification stack for onboarding and KYC flows. It verifies a real person holding a real document, then returns a structured decision:

Document checks: front/back ID capture with authenticity scoring.
Liveness: short selfie video with anti-spoofing.
KYC endpoints: hosted at https://api.sagea.space/helios/kyc/v1 and /v2.

Open Helios ↗

How Helios works

Capture, verify, decide — authenticity, face match, and liveness.

Run a KYC request

Front/back images plus liveness video in 10 minutes.

Studio

Studio is the hosted model platform. Generate API keys, test prompts in the Playground, and call chat, audio, and OCR models from a single base URL:

https://api.sagea.space/v1

Open Foundry console ↗

Studio overview

Keys, Playground, and SDKs.

Send your first API request

Key → install → chat completion.

API Reference

Full endpoint reference.

MYLO

MYLO is the voice agent layer on top of Studio. Connect Sonus TTS + SAGE reasoning + telephony/WebSocket transport to ship real-time agents:

Sessions: realtime agent instances over WebSocket with ~200ms first byte.
Calls: outbound appointment reminders and inbound IVR replacement.
Transcripts: turns, recordings, and analytics per session.

MYLO overview

Sessions, calls, transcripts.

Launch your first agent

Realtime session in 10 minutes.

RUNE

RUNE is the local-first coding agent. It runs SAGE reasoning models on your machine through Ollama — explore, edit, and test code with natural language:

Private by default: files and prompts never leave your hardware.
Context-aware: automatic file-tree and Git-status scanning.
Approval-based: default, plan, accept-edits, and auto-approve profiles.

GitHub repo ↗

Install RUNE

One-line install plus first model pull.

Scaffold a project

Full project from a single prompt.

Foundry

Foundry is the on-prem distribution of SAGEA. Same models, your VPC — no hosted APIs, no shared keys.

Private deployment of SAGE, Sonus, and ARVA weights
Your networking, logging, and data retention
Commercial license + support contract

Contact sales ↗

Foundry overview

Private VPC deployments.

Studio overview

Prefer hosted APIs instead.

Common questions

Where do I get an API key?

Helios and Studio use hosted keys from Foundry: API keys → Create new key, then export SAGEA_API_KEY="...". Foundry on-prem uses license-based auth instead — see Foundry.

Which product should I start with?

KYC / onboarding → Helios
Chat / voice / docs via API → Studio
Real-time voice agents → MYLO
Coding help on your machine → RUNE
Private VPC / regulated data → Foundry

What's the difference between Studio and MYLO?

Studio is the raw model APIs (chat, speech, OCR) you call yourself. MYLO is the managed voice-agent layer on top: sessions, phone calls, transcripts, and realtime audio transport — use it when you want conversational agents without building the voice pipeline.

What's the difference between Foundry and Studio?

Studio is hosted: your requests go to api.sagea.space with an API key. Foundry is on-prem: the same model weights run inside your VPC under a commercial license, and your data never leaves your infrastructure.

Can I run everything offline?

RUNE, yes — models, files, and prompts stay on your machine. Helios, Studio, and MYLO are hosted APIs and need connectivity, except for Foundry on-prem deployments.

On this page