Overview
AnyInt is not only a transport layer. The platform also changes how applications behave once they are in production: how output is streamed, how keys are limited, how routing is applied, and how model output is shaped for downstream systems.
AnyInt is not only a transport layer. The platform also changes how applications behave once they are in production: how output is streamed, how keys are limited, how routing is applied, and how model output is shaped for downstream systems.
Feature map
| Feature area | What it affects |
|---|---|
| Streaming | how fast users see output |
| Structured outputs | how safely you can feed model output into code |
| Tool calling | how models interact with external systems |
| Prompt caching | how repeated context can reduce cost and latency |
| Routing and fallbacks | how traffic moves across models and providers |
| Guardrails | how keys, teams, and policies control usage |
| Integration verification | how to prove auth, model IDs, streaming, async tasks, callbacks, and retries work before launch |
Some of these are model-dependent. Others are account, key, or organization guides. Use the linked pages below to decide where a behavior actually lives.
How to read this section
- read Streaming if your UI depends on low-latency output
- read Structured Outputs and Tool Calling if model output feeds code or workflows
- read Smart Routing, Models, and Model Fallbacks if reliability and cost control matter
- read Verify Your Integration before shipping a production customer integration
Errors and Limits
AnyInt publishes both bearer-style and provider-style auth patterns. If a request fails unexpectedly, verify that you used the right header family for the route you called.
Streaming
Streaming is the simplest way to improve perceived latency in chat, copilots, and interactive generation flows.