gpt-5.6-sol

OpenAIgpt-5.6-sol

GPT-5.6 Sol is the flagship model in OpenAI's GPT-5.6 series. It is suited for complex reasoning, coding, and agentic workflows, and is particularly strong at command-line and multi-step coding tasks and long-horizon problem solving.

API Pricing

Price typeBilling expression
Official price
Base pricingPricing tier: base
  • Input usage × $4.00
  • Output usage × $20.00
  • Cached read × $0.40
  • Cache creation × $5.00
GPT_Pro×0.660% of official list price
Base pricingPricing tier: base
  • Input usage × $2.40
  • Output usage × $12.00
  • Cached read × $0.24
  • Cache creation × $3.00

Group multiplier: x0.6

GPT_Ultra×0.880% of official list price
Base pricingPricing tier: base
  • Input usage × $3.20
  • Output usage × $16.00
  • Cached read × $0.32
  • Cache creation × $4.00

Group multiplier: x0.8

Official pricing is subject to the model provider's latest published pricing and may vary by request method or service region. GPT_Pro, GPT_Ultra — TokenByte channel pricing. Actual prices and availability reflect current platform data.

Model information

Model providerOpenAI
Model IDgpt-5.6-sol
Context window1050000
InputTextImage
OutputText

Frequently asked questions

01

What is GPT-5.6 Sol?

GPT-5.6 Sol is the flagship model in OpenAI's GPT-5.6 series. It is suited for complex reasoning, coding, and agentic workflows, and is particularly strong at command-line and multi-step coding tasks and long-horizon problem solving.
02

How much does GPT-5.6 Sol cost?

GPT-5.6 Sol costs $2.00/M input tokens and $10.00/M output tokens, with separate rates for Cache Read at $0.20/M tokens, Cache Write at $2.50/M tokens, and Web Search at $10.00/1K calls.
03

What is the context length of GPT-5.6 Sol?

GPT-5.6 Sol has a 1,050,000 token context window. It supports up to 128,000 completion tokens.
04

Does GPT-5.6 Sol support tool calling and structured outputs?

Yes. GPT-5.6 Sol accepts `tools` and `tool_choice` for function calling. It also supports structured outputs via a JSON schema in `response_format`.