Developers Guide

IuVe is a private AI workspace with API access, permissioned GitHub integration, and production-ready chat workflows. Use IuVe programmatically via REST API or embed it in your applications.

🔌 REST API Overview

IuVe exposes a REST API for programmatic access to chat, batch operations, notes, API key management, and usage tracking.

Base URL: https://www.iuve.eu/api

Method	Endpoint	Purpose
POST	/chat	Send a single message
POST	/chat/stream	Stream a response
POST	/chat/batch	Batch multiple requests
GET	/usage	View usage stats
GET	/api-keys	Manage API keys
GET	/github/repos	List connected repos

🔑 Authentication API Key

Generate API keys from your account dashboard. Include the key in a header:

Authorization: Bearer YOUR_API_KEY

Each key can be rate-limited and scoped. Revoke keys anytime from the account settings.

💬 Chat Workflows

/api/chat — Single request-response chat:

POST /api/chat
{
  "prompt": "Analyze this code",
  "mode": "chat",
  "files": []
}

/api/chat/stream — Streaming response (SSE):

POST /api/chat/stream
# Response: text/event-stream with chunks

/api/chat/batch — Process 5-50 requests async:

POST /api/chat/batch
{
  "requests": [{...}, {...}],
  "callback_url": "https://your-domain/webhook"
}

📦 File and Repository Access

Upload files (up to 10 per request, 25 MB each) or connect GitHub with explicit consent.

Files: multipart/form-data with files

GitHub: OAuth grant + consent for repo/code access. Retrieve repos, browse files, propose changes.

Web Agent: Crawl and analyze website structure for research or embedding workflows.

📊 Usage and Billing

GET /api/usage — Summary by day/week/month

Track input characters, output characters, and request counts. Usage limits depend on your plan.

Plans: Free (limited), Starter, Pro, Team. Upgrade in account settings.

🛠️ Examples and SDKs

Python:

from iuve_client import IuVeClient
    client = IuVeClient(api_key="sk-...")
resp = client.chat("What is this code?
code...")
print(resp.text)

JavaScript:

import { IuVeClient } from '@iuve/client';
    const client = new IuVeClient({apiKey: 'sk-...'});
const resp = await client.chat({prompt: '...'});
console.log(resp.text);

Full examples and SDKs: https://www.iuve.eu/developers/sdk

⚙️ AI Tools and Modes

Chat mode: General conversation and decision-making.

Research mode: Web search, deeper analysis, structured output.

Coding mode: Terminal, IDE review, repository analysis, PR proposals.

Website Agent: Crawl, analyze, and embed website knowledge.

Pass mode in the request: "mode": "research" or "mode": "coding"

🚀 Latest IuVe AI Achievements

SSE streaming: production token streaming via /chat/stream.
Batch API: multi-request execution with per-item status via /chat/batch.
Cache layer: similar-request cache with TTL and similarity threshold.
Adaptive limits: plan and tenant-aware throttling policy.
Security hardening: injection guardrails, audit events, and alert thresholds.
Governance gates: quality, safety, and checksum validation before fine-tune promotion.

🔒 Security and Rate Limits

Rate Limits: Default 60 requests/min per account. Higher plans include additional limits.

HTTPS only. No persistent logs of chat output unless explicitly saved.

Secrets and sensitive data: Avoid submitting API keys, passwords, or PII unless your workflow is specifically designed for it.

See our Acceptable Use Policy and Privacy Policy for details.

❓ Frequently Asked

Can I use IuVe for production? Yes. IuVe is built for production workloads with API access, team collaboration, billing, and admin controls.

What if I hit a rate limit? Upgrade your plan or wait for the rate-limit window to reset (typically 1 minute).

Can I export my data? Yes. Use the account page to request a data export. Billing records are available anytime.

Is there SLA or uptime guarantee? IuVe runs on stable infrastructure. See our Terms of Use for limitations.