Token Market is an LLM gateway for developers and enterprise teams. It brings three jobs into one workflow: finding the right model, calling models reliably, and managing API costs over time. Developers use one unified API to access multiple model capabilities, while the platform provides Model Market, provider data, Smart Routing, workspace governance, and usage monitoring.
A single Token Market integration gives AI applications Smart Routing, enterprise-grade reliability, and privacy-first security across text, image, video, and audio models. We do not log, store, or use your prompts or generated content to train models. Model and provider capabilities, pricing, and request limits change with the live catalog and upstream configuration. Treat the model details page, console, and current API documentation as the source of truth.
The platform supports two related workflows. For model selection, teams can compare models and providers by pricing, latency, throughput, reliability, and API capabilities. For model access, a unified endpoint and routing policies reduce duplicate integrations, so application code does not depend on a single provider.
Open a model in Model Market to review its basic information, supported API capabilities, provider pricing, performance data, and unified API examples. Provider data supports model selection, while continuous collection and testing produce the evaluation metrics. Results vary by model, provider, time window, and request conditions. See “Data collection, evaluation, and presentation” under Performance metrics for the methodology.
The platform provides one API endpoint: https://api.tokensmarket.ai/v1. Use an API Key from the current workspace and include the model ID in each request. For chat, Embedding, Reranker, image, video, and audio capabilities, follow the current examples in the relevant API documentation and model details page.
A unified API does more than replace a URL. Model changes, provider selection, and supported routing preferences can stay in platform configuration or request parameters, while application code keeps a stable input and output flow. Before integrating, read API-Key and the relevant API examples.
A model may be available through multiple provider channels. On the model details page, compare providers by pricing, latency, throughput, availability, and other criteria, or let Smart Routing select a channel under the current policy.
Individual developers and enterprise teams manage API resources through workspaces. Each API Key belongs to and receives permissions from its current workspace. You can configure expiration, quota, allowed models, token groups, and an IP allowlist. Lists show only masked values, and the platform reveals the complete Key only once, immediately after creation.
Enterprise teams can combine business verification, workspace membership, and role-based permissions to separate collaboration from API resources. Always store each API Key on the server. Never place it in browser code, repositories, logs, or public support tickets. If a Key leaks, disable or delete it immediately and create a replacement.
The console records the model, provider, token usage, duration, status, and cost for each request. Teams can use these details to investigate failures, compare routing behavior, and analyze costs. Current pages and request results determine model pricing, workspace quota, and final charges. A request may still incur charges after the client disconnects if the upstream provider has already started processing it.
Enterprise users can link a verified business entity and manage workspace members, roles, permissions, and API resources. The console combines verification, access control, API Key management, request logs, and billing so teams can collaborate by responsibility and trace issues. Available menus and actions depend on account status, workspace, and permissions.
