Overview
OpenAI API is a developer platform that exposes OpenAI’s large language-and-multimodal models—such as GPT-4o, GPT-4, GPT-3.5-turbo, DALL·E, and Whisper—through a set of RESTful endpoints.
Key features
- Unified interface: same request/response schema across model families.
- Multi-task capability: text and code generation, summarization, translation, extraction, classification, image creation/editing, speech-to-text, and more.
- Fine-tuning & embeddings: customize base models on domain data and obtain vector representations for semantic search.
- Function & tool calling: safely connect model outputs to external functions or tools.
- Security & governance: per-account rate limits, user-ID tracking, content-filtering and moderation endpoints.
- Official SDKs: maintained Python and JavaScript/TypeScript clients plus community libraries for many other languages.
Pricing & availability
Usage-based, per-token or per-asset pricing with enterprise tiers for higher throughput, dedicated capacity, and data-retention controls. Available globally except in restricted regions.
Typical use-cases
- Conversational assistants and chatbots
- Retrieval-augmented generation (RAG) search systems
- Code completion and review tools
- Content moderation workflows
- Creative media generation (stories, marketing copy, images, audio)
Governance
OpenAI enforces policy-based safeguards, provides model output filtering, and publishes detailed policy documents to help developers build responsibly.
