Unified infrastructure
for developers and AI agents.
From coding agents such as Claude Code and Codex to production AI applications, unify model access, routing, authentication, billing and monitoring.
# Claude Code
export ANTHROPIC_BASE_URL=https://api.nexinfer.com
export ANTHROPIC_AUTH_TOKEN=sk-nexinfer-xxxx
claude
# Codex / OpenAI-compatible clients
export OPENAI_BASE_URL=https://api.nexinfer.com/v1Compatible endpoints for developer tools
Built around the things that matter in production.
Clear privacy boundaries, stable system operations, and usage-based pricing for developers and AI agents.
Security and privacy by boundary
nexinfer does not use request content for model training and controls access through account and API key permissions. Content is processed transiently for inference; only essential operational metadata is retained.
Stable system operations
Health checks, service monitoring and rollback-ready releases support consistent gateway and dashboard operation, with failures traceable and changes recoverable.
Cost-efficient by design
Compare current model prices in the catalog and pay against recorded usage instead of relying on stale marketing prices.
Change the base URL, keep your workflow.
Use the key generated in your dashboard with clients that support a custom Anthropic or OpenAI-compatible endpoint.
- Separate Anthropic-compatible and OpenAI-compatible endpoints
- Streaming and tool-call capable gateway routes
- Current model availability is shown in the dashboard
# Claude Code
export ANTHROPIC_BASE_URL=https://api.nexinfer.com
export ANTHROPIC_AUTH_TOKEN=sk-nexinfer-xxxx
claudeUse an API key created in your nexinfer dashboard. Availability and account limits can change.
What is available today
What is nexinfer? +
nexinfer is a unified infrastructure platform for developers and AI agents. Today it provides model access, routing, authentication, billing and monitoring through a developer dashboard and compatible endpoints for Claude Code, Codex and other clients.
Does it work with Claude Code and Codex? +
Both tools can use custom API endpoints. Follow the nexinfer Quickstart for the current environment variables and base URLs.
Which models are available? +
Availability changes with model-provider capacity and account permissions. Check the model catalog for current models, prices, and group multipliers.
What data is recorded for usage? +
nexinfer does not use request content for model training. The service records operational metadata needed to run, secure and meter requests, including model, token counts, timing and status information. Model-provider policies still apply; see the Privacy Policy for details.
Where are prices and payment methods listed? +
nexinfer does not currently publish a marketing price or payment-method claim. A public pricing page will be added only after rates and checkout are verified against production billing.
Is nexinfer an OpenRouter alternative? +
It can be an alternative for teams that primarily need custom Anthropic-compatible and OpenAI-compatible endpoints. OpenRouter remains a better fit when its broader routing marketplace is the main requirement.
One infrastructure layer for developers and AI agents.
Create an account, generate a key and follow the Quickstart. Current availability and account limits are shown in the dashboard.