Live provider/model prices, per 1,000,000 tokens (USD). Refreshed daily from provider APIs. Cached locally — see website/aio/data/prices.current.json.
| Provider | Model | Input $/1M | Output $/1M | Currency |
|---|---|---|---|---|
| Loading… | ||||
See which model the live router would pick across your providers for a given strategy/tier. Uses the reference aio-router (mirrors the Rust engine).
Cheapest models by blended price (input+output)/2 across all providers. Filter by capability keyword.
Per-model request counts from the live usage log. Updates as requests flow through the gateway.
| Provider / Model | Requests |
|---|---|
| Loading… | |
GPU providers available to you for running LLM inference.
AI models available across your providers.
LXD container + service topology.
Board grouped by lifecycle status. Each column has an editable title + prompt that tunes the work of its cards. The design plan tells the story and tracks it all. Empty columns collapse to a thin strip — hover or click to expand.
/home/dev/git/aio/modules — crates (blue), internal modules, and use aio_* cross-crate edges./home/dev/git/aio/modules/aio-neural source./home/dev/git/aio/modules/aio-reasoning source./home/dev/git/wiki — categories and pages.127.0.0.1:8765)./home/dev/git/aio/modules/aio-core).~/.hermes/memory_store.db) — facts, entities, and memory banks./api/prices/optimize and /api/router/simulate endpoints.~/.hermes/skills) and plugins (~/.hermes/plugins)./home/dev/git/aio/docs/design/*.md./home/dev/git/aio/.dashboard-settings.json (no secrets)./api/runtime + gateway capacity alerts./api/agent/edge/backups).Your personal token-spending budget. Admins can adjust limits in the Admin Panel.
Live system health, API metrics, and the LXD edge-node fleet. Admin only.