FluxionAPI Docs

Theme

About Fluxion AI ​

Fluxion AI is a unified platform for accessing and managing multiple AI APIs. With one account and a single interface, users can access currently available leading AI models—including GPT, Claude, Grok, DeepSeek, GLM, and Kimi—and centrally manage API keys, model routes, usage, and costs.

The platform brings fragmented model APIs, invocation protocols, route selection, and billing records into one service, reducing the overhead of maintaining multiple accounts, integrations, and management systems. Whether you use AI coding tools, clients that support custom endpoints, or SDKs and APIs integrated into existing applications, you can work with one consistent system for accounts, keys, and usage management.

Unified Access to Leading AI Models ​

Model providers typically use different account systems, API protocols, and billing methods. Fluxion AI provides one model gateway, bringing model selection and route management together on a single platform.

  • Multi-model access: Access GPT, Claude, and other currently available text, reasoning, and image models.
  • Unified API key management: Create and manage API keys for different projects, models, and billing groups within one account.
  • Compatible APIs: Use OpenAI-compatible, Anthropic-compatible, and related model APIs to reduce migration work for existing tools and applications.
  • Public Model Plaza: Review model identifiers, base prices, available routes, group multipliers, and update times in one place.

Models and routes may change based on actual platform availability and operating conditions. Refer to the Model Plaza for the current selection.

Multi-Route Scheduling and Request Management ​

Fluxion AI provides more than model access. It manages the complete lifecycle of a model request, from initial access through usage recording.

  1. Authentication and key validation: Validate API key status, assigned group, account permissions, and balance or subscription status.
  2. Model and route selection: Determine the request path from the model identifier, API protocol, group rules, and current route configuration.
  3. Available resource selection: Select from eligible routes and resources to reduce the effect of a single route fluctuation on continued use.
  4. Protocol adaptation and model invocation: Forward requests and return responses according to APIs such as OpenAI, Responses, Images, or Anthropic Messages.
  5. Usage and cost metering: Calculate costs from the model's base price, actual token usage or request count, route multiplier, and plan rules.
  6. Usage records and analytics: Record the necessary request metadata and provide detailed and summarized usage data in the dashboard.

Route scheduling, retry behavior, and compatibility depend on the platform's current implementation and published documentation. They should not be treated as an unverified guarantee of absolute availability.

Transparent Model, Pricing, and Usage Information ​

Fluxion AI separates information for selection, invocation, and reconciliation between public pages and the signed-in dashboard, making each part of the process easy to verify.

Models and Pricing ​

The Model Plaza publishes model identifiers, official or base prices, available groups, route multipliers, and billing methods. Users can confirm model and route information before creating an API key or sending a request.

Model IQ Monitoring ​

Model IQ regularly records IQ scores, passed tests, and trends for selected models using a fixed set of evaluation tasks. These results can inform model selection and anomaly investigation, but reflect only the specified test set and evaluation time and do not replace validation for your own use case.

Official Service Status ​

Official Service Status summarizes operating status and incidents published by services such as OpenAI and Anthropic, helping users distinguish provider-side disruptions from local configuration issues. This page reports official provider status and is not an SLA for Fluxion AI's own routes.

Usage and Cost Details ​

After signing in to the dashboard, users can review request counts, token usage, actual costs, success rates, latency, model and group distribution, and per-request model, endpoint, token, and charge details. Full filtering, failed-request lookup, and data export are available in the dashboard.

Flexible Billing and Service Options ​

Fluxion AI supports two primary usage options: a pay-as-you-go balance for usage-based billing and subscription plans that provide allowances and group rules for ongoing use. Model base prices, route multipliers, plan discounts, and availability may change with service configuration. Check the purchase page and Model Plaza for current information before completing a purchase.

The platform also supports RMB settlement, personal or business invoicing, Chinese-language technical support, and consultation for team usage or dedicated resources. The dashboard and latest service documentation define the current scope and process for these services.

Get Started ​