GPT-5.6 Sol

LLM Provider / Model · Frontier coding, agentic work and complex professional tasks

At a glance

Input price	$5
Output price	$30
Cache price	$0.50
Price tier	Premium
Context window	1M+
Max output	128K
Context tier	500K+
Speed tier	Slow/Reasoning
Latency	high
Knowledge cutoff	Feb 2026
Modality	Multimodal (vision)
Model ID	gpt-5.6-sol
Provider	OpenAI
Benchmark	OpenAI’s frontier tier; SWE-bench no longer reported
Released	Jul 2026
Hosting	Closed/API
Capabilities	Vision, Tool use, Reasoning effort control (none-max), Fast mode, Streaming, Structured output, Prompt caching

Vision, Tool use, Reasoning effort control (none-max), Fast mode, Streaming, Structured output, Prompt caching

Frontier coding, agentic work and complex professional tasks

Kiro's spec mode is tuned for Claude models — GPT-5.6 Sol may work but spec-driven flows are best with Claude.
Gemini CLI is locked to Google Gemini models — your LLM is GPT-5.6 Sol.
Claude Code only works with Claude models — your LLM is GPT-5.6 Sol.
Devin uses Cognition's proprietary model internally — your selected GPT-5.6 Sol will be ignored.
OpenAI Codex (Cloud) runs GPT-5.1 Codex internally — your selected GPT-5.6 Sol will be ignored.
Jules runs Gemini models internally — your selected GPT-5.6 Sol will be ignored.
The Claude Code SDK is tuned for Claude models — GPT-5.6 Sol may work via the model adapter but loses Claude-specific features (thinking, prompt caching, sub-agents).
Lovable uses a fixed internal model — your selected GPT-5.6 Sol will be ignored.
Replit Agent 3 uses its own model selection internally — your selected GPT-5.6 Sol will be ignored.

Build a full stack around GPT-5.6 Sol — Flowpicker shows compatibility warnings before you commit.