AI MODEL API

OpenAI API

Provides hosted models and tools that applications can call for text, reasoning, search, image, audio, and agent workflows.

Last verified September 10, 2026

Best for / Not for

BEST FOR

  • Products that need model capabilities through an API
  • Teams that want multiple model tiers behind one platform
  • Applications with measurable per-request AI workloads

NOT FOR

  • Products that require fully self-hosted model inference
  • Workloads without controls for token and tool usage
  • Use cases that cannot send data to an external model service

Where it fits in your stack

Connects server-side product workflows and retrieved application data to hosted AI models, with results returned to the interface. It replaces self-hosted inference for supported workloads but does not replace the application backend or its access controls.

Pricing

Pricing model: Pure usage-based pricing with no fixed monthly fee. Token prices (input / output, per 1M tokens): GPT-6 Astra $10 / $50 · GPT-5.6 Sol $4 / $20 (promotional; regular $5 / $30) · GPT-5.6 Terra $2 / $12 · GPT-5.6 Luna $0.20 / $1.20. Batch API is 50% less. Data residency requirements add 10%. The web search tool is $10 / 1,000 calls. Prices apply to contexts below 272K.

Free allowance: not verified

Overage rule: not verified

Trade-offs

  • The promotional price for GPT-5.6 Sol runs through at least 21 November 2026. After that, input could rise 25% and output 50%.
  • Cost varies with model choice, context size, output length, and tools
  • Model behavior and available capabilities can change between releases

Sources

Last verified: 2026-09-10