Standard Compute
← All models

GPT-5.6 Sol

Frontier

OpenAI's new flagship tier (July 2026) — built for the hardest coding and research problems.

OpenAI
Maker
95/100
Overall
44
Value score
$13
Blended /1M
1.1M
Context
Proprietary
Weights
Pricing: $5.00 in · $30 out / 1M tokens · synced from OpenRouter · updated 2026-07-17

Scorecard

Output Quality9.7/10 · 81% community
Agentic Ability9.6/10 · 80% community
Speed7.4/10 · 70% community
Value for $7.8/10 · 71% community
Reliability9.3/10 · 79% community

GPT-5.6 Sol is the flagship of OpenAI's new three-tier family (released July 9, 2026), aimed at the hardest problems: complex coding, security research, long agentic runs. In the new naming scheme the number is the generation and Sol/Terra/Luna are durable capability tiers. Sol pairs a ~1M-token context with $5/$30-per-million pricing, undercutting Anthropic's premium tier while competing at the frontier.

Strengths
  • Frontier-class quality at lower price than Anthropic's premium tier
  • ~1M-token context
  • Strong long-run agentic behavior
Trade-offs
  • Very new — real-world agent track record still forming
  • Slower and pricier than Terra for routine work

Agents that run Sol well

Hard coding and research tasks for teams in the OpenAI ecosystem.

Best model for CodexBest model for ClineBest model for AiderBest model for OpenCodeBest model for CursorBest model for Kilo Code

GPT-5.6 Sol FAQ

Is GPT-5.6 Sol good for AI agents?
Hard coding and research tasks for teams in the OpenAI ecosystem. On our editorial scoring it rates 9.6/10 for agentic ability and 9.7/10 for output quality. Frontier-class quality at lower price than Anthropic's premium tier.
How much does GPT-5.6 Sol cost?
Via OpenRouter, GPT-5.6 Sol is priced at $5.00 in · $30 out / 1M tokens — a blended rate of about $13 per 1M tokens for typical input-heavy agent use.
What is GPT-5.6 Sol's context window?
GPT-5.6 Sol has a 1.1M-token context window — large enough to hold sizeable codebases or document sets in a single run. OpenAI is the maker.
Can I self-host GPT-5.6 Sol?
No — GPT-5.6 Sol is a proprietary, hosted-only model from OpenAI. You access it through an API (e.g. OpenRouter) rather than running the weights yourself.
What are the downsides of GPT-5.6 Sol?
The main trade-offs: very new — real-world agent track record still forming; slower and pricier than terra for routine work. It's strongest for hard coding and research tasks for teams in the openai ecosystem.

Sol compared

Sol vs Fable 5Sol vs Opus 4.8Sol vs GPT-5.5Sol vs Gemini 3.1 ProSol vs GLM 5.2Sol vs Qwen3.7 MaxSol vs DeepSeek V4 ProSol vs Kimi K2.6

The model is half the story — the agent is the other half

The model picks the moves; the agent runs the loop, the tools, and the guardrails. Once you've chosen a model, see which agent gets the most out of it.

Compare AI agents →