Getting Started with DeepToken

Welcome to DeepToken

DeepToken is a multi-model inference gateway and billing control plane. One OpenAI-compatible API lets you route requests across multiple providers (OpenAI, Anthropic, DeepSeek, Qwen, Gemini, Llama and more) with built-in fallback, transparent token metering, organization-level billing, and platform-grade observability.

Creating Your Account

1. Sign up: Visit the signup page and create your account with your email or via Google, Microsoft, or Apple.
2. Pick a plan: Choose Free to start or any paid tier for more monthly tokens, higher request rate limits, premium model access, and longer log retention.
3. Verify your email: Click the verification link to activate your account.
4. Create an organization (optional): If you work with a team, create an organization, invite members, and assign roles before issuing shared API keys.

Issuing Your First API Key

1. Log in at deeptoken.app and open the dashboard.
2. Open API Keys and click Create Key. Scope the key to specific models, a monthly token budget, and an optional IP allow list.
3. Copy the secret immediately — it is shown once and stored only as a hash on the server.

Making Your First Request

DeepToken implements the OpenAI Chat Completions schema. Point any OpenAI-compatible SDK at the gateway:

• Base URL: https://api.deeptoken.app/v1
• Authorization header: Bearer dt-... (your API key)
• Model field: any model id from the catalog, for example gpt-4o, claude-3-5-sonnet, deepseek-r1, qwen2.5-72b-instruct, llama-3.3-70b-instruct, or gemini-2.0-flash.

The gateway will resolve the request to the configured upstream provider, apply your routing and fallback rules, meter token usage, and stream the response back through the same connection.

Routing and Fallback

1. Open Models and pick the model you want to configure. Each model can have a primary provider and an ordered fallback chain.
2. Set policy: define timeouts, retry limits, and the health threshold that triggers fallback.
3. Monitor: watch the dashboard for fallback rates and provider health, then refine as traffic grows.

Usage, Billing, and Observability

• Usage dashboard: real-time token, request, and cost breakdown by API key, model, and route.
• Request log: per-request latency, status, retries, fallback hops, and token counts.
• Audit log: append-only record of routing, key, member, and billing changes.
• Invoices and exports: monthly invoice plus CSV export for finance reconciliation.

Next Steps

• View Pricing
• Explore Platform Capabilities
• Frequently Asked Questions