> ## Documentation Index
> Fetch the complete documentation index at: https://developers.t2000.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Console

> agents.t2000.ai — the agent directory and the console behind it. Sign in with Google (zkLogin) to manage API keys, billing, usage, and your agents — same Passport + balance as Audric.

**agents.t2000.ai** is **t2 Agents**: skills, the public **[directory](https://agents.t2000.ai/agents)** (every registered [Agent ID](/agent-id), no login), and the **console** at **[/manage](https://agents.t2000.ai/manage)** — keys, credit, usage, and your **agents**, under one account.

You sign in with **Google (zkLogin)**, which is the **same Passport identity + credit balance** as [Audric](https://audric.ai) and `api.t2000.ai`. One account spans the consumer app, the API, and your agents — no second login, no separate billing.

<Note>
  The console is the **human** surface (point-and-click, zkLogin). The **`t2` CLI** is the **agent/programmatic** surface (keypair-signed). Same account, same keys, same credit ledger.
</Note>

***

## Sign in

Go to [agents.t2000.ai/manage](https://agents.t2000.ai/manage) → **Sign in with Google**. A non-custodial Passport (zkLogin wallet) is created/loaded deterministically — the same address you'd get on `audric.ai`.

***

## API keys

**API keys → Create API key.** The secret (`sk-…`) is shown **once** — store it safely. Keys authenticate apps + agents to [Private Inference](/private-api). Revoke (and recreate) any key that leaks.

***

## Billing — credit and wallet, two pots

**Credit** (USD-denominated) pays for API usage, funded either way:

* **Card** — Stripe top-up + optional **auto-recharge**.
* **USDC / USDsui** — pay from your Passport, gasless (**Billing → Pay with stablecoin**).
* **Plans** — an Audric Pro/Max plan's included credit spends here too.

**Wallet USDC** (on-chain, shown on the dashboard with your deposit address) is the second pot: it funds **x402 payments** (`t2 pay`) and anything else the agent spends on-chain. Credit pays for tokens; the wallet pays the world.

***

## Usage

**Usage** shows tokens + spend by model and request counts — your real consumption, per model.

***

## Models

**Models** is the live catalog with per-model pricing, context window, and **privacy tier** (Private / Confidential) — the same data `GET /v1/models` returns.

***

## Your agents

Everything agent-side lives under **My agents**:

* **New agent** — the console shows the `t2 agent create --owner <your-passport>` command to run where the agent lives (its key never touches the browser); the ownership request then appears here for a one-tap confirm.
* **Become an agent** — one tap on **Create your Agent ID** registers your Passport in the on-chain registry (sponsored, consent-first: it lists your address in the public directory; deactivate anytime).
* **The agent you are** — because your Passport *is* this agent, you can edit its profile directly from the browser, zkLogin-signed.
* **Agents you own** — edit profiles and manage the on-chain active flag. Ownership confirmations are one gasless signature (an agent proposes you with `t2 agent link`; you confirm here — and can unlink later).

***

## Where next

<CardGroup cols={2}>
  <Card title="Private Inference" icon="lock" href="/private-api">
    The OpenAI-compatible endpoint this console manages — private by default, verifiably confidential when it matters.
  </Card>

  <Card title="Agent ID" icon="fingerprint" href="/agent-id">
    Identity, handles, ownership, and the public directory.
  </Card>
</CardGroup>