AI API glossary
Definitions and practical decisions for teams building model-powered applications.
OpenAI-compatible API
Understand OpenAI-compatible APIs, endpoint support and SDK integration. Learn what to test before connecting an existing chat application to a model service.
Updated 1 Nov 2026
India-hosted inference
Learn what India-hosted inference means for an AI application. Review model routes, request processing and the data flows beyond the model server itself.
Updated 1 Nov 2026
DPDP Act and AI inference
Understand personal data in AI inference workflows. Map application data flows and review DPDP-ready controls alongside the design of your complete application.
Updated 1 Nov 2026
Open-weight models
Learn what open-weight models provide, how licences affect their use and what to evaluate when choosing between running weights and using a managed model API.
Updated 1 Nov 2026
Tokens and token pricing
Understand input and output tokens, per-million token rates and workload estimates. Learn how conversation history and retrieved passages affect AI usage costs.
Updated 1 Nov 2026
Context window
Understand context windows, token budgets and served model limits. Learn how to reserve answer space and evaluate long-document requests in your application.
Updated 1 Nov 2026
Tool calling
Learn how tool calling connects a language model to application functions. Follow the request cycle and understand validation, permissions and execution.
Updated 1 Nov 2026
Embeddings and rerankers
Understand embeddings and rerankers in document search. Learn how candidate retrieval and relevance scoring work together, and how to evaluate answers.
Updated 1 Nov 2026