OpenAI Responses API and Agents SDK

New tools for building agents and tools

Website platform.openai.com

What it is

GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model. It accepts both text and image inputs, and produces text outputs (including Structured Outputs). It is the best model for most tasks, and is our most capable model outside of our o-series models.

Intent

I need it when

Reduce latency and costs when processing high volumes of API requests

GPT-4o provides optimized inference performance and token efficiency compared to earlier models, lowering per-request costs and response times at scale

Access state-of-the-art multimodal AI capabilities for vision and text tasks

GPT-4o natively processes images, text, and other modalities in a single model, enabling developers to build applications that understand and reason across multiple input types

Integrate structured AI responses into applications with predictable output formats

The Responses API allows developers to define and enforce response schemas, ensuring AI outputs conform to required data structures for reliable downstream processing

Build AI agents that can autonomously complete multi-step tasks and make decisions

The Agents SDK enables developers to create autonomous agents using GPT-4o that can plan, execute, and iterate on complex workflows without manual intervention at each step

Drop

Not a fit when

User requires on-premise or self-hosted deployment with no external API calls
Organization has strict data residency requirements prohibiting cloud-based AI processing
User needs real-time latency under 100ms for mission-critical applications
Budget constraints prohibit per-token or usage-based pricing models
User requires guaranteed SLA uptime above 99.99% with contractual penalties

Commercials

Pricing

Pricing not specified