Model
Model explorer

GPT-5.6 Sol

CLOSED
OpenAI · GPT-5.6 family · released Jul 9, 2026

Flagship (Sol) tier of the three-tier GPT-5.6 family, billed as OpenAI's best coding model yet (Terminal-Bench 2.1 SOTA); limited-preview June 26 under US govt review, GA July 9, 2026. Sibling tiers GPT-5.6 Terra and GPT-5.6 Luna are tracked as their own corpus models in the same family (OpenAI: 'durable capability tiers that can advance on their own cadence').

ReasoningCodingVisionFunction callingTool useAgentic
2948.0
Elo · rank #2
Parameters
Undisclosed
Active params
Context
1.05M tokens
Architecture
Undisclosed architecture (presumed dense transformer); flagship tier of the three durable capability tiers Sol/Terra/Luna, each tracked as its own corpus model (like Claude Opus/Sonnet/Haiku)
License
Proprietary
Languages
API price (in/out)
$5 / $30
Modalities
text · vision
Benchmark results
Bar shows position within the tracked field; marker = field best
Agents' Last ExamAgents52.7%#1
best: this model · 52.7%
ARC-AGI-1Reasoning96.5%#1
best: this model · 96.5%
ARC-AGI-2Reasoning92.5%#1
best: this model · 92.5%
BrowseCompAgents90.4%#2
best: Kimi K3 · 91.2%
CyberGymCoding84.5%#1
best: this model · 84.5%
FrontierBenchAgents34.4%#1
best: this model · 34.4%
best: this model · 89.0%
best: this model · 83.0%
GDPval-AAAgents1747.8#4
best: Claude Fable 5 · 1932
GPQA DiamondReasoning94.6%#1
best: this model · 94.6%
Humanity's Last ExamReasoning47.2%#9
best: Claude Sonnet 5 · 57.4%
IFBenchReasoning72.7%#21
best: MiniMax M3 · 83.0%
MMMU-ProVision83.0%#4
best: Claude Opus 4.7 · 85.5%
SWE-bench ProCoding64.6%#4
best: Claude Fable 5 · 80.0%
τ³-BankingAgents33.0%#1
best: this model · 33.0%
Terminal-Bench 2.0Coding88.8%#1
best: this model · 88.8%
Run it locally
Closed weights — available via API only. No local deployment.
Input / M tok
$5
Output / M tok
$30
API price $5/$30 · each benchmark row carries its own source badge (see methodology)