Back

OpenAI

GPT-5.6 Sol

ReasoningClosed weights
Intelligence
42.3
Rank 49 of 744 measured models
Coding
77.2
Rank 7 of 298 measured models
Output speed
81 t/s
Rank 228 of 403 measured models
Price/1M
$8
Rank 603 of 659 measured models
Context window
1M
Rank 84 of 1199 measured models
Cost per task
$0.808
Rank 119 of 193 measured models

GPT-5.6 Sol (High) scores 42.3 on the Artificial Analysis Intelligence Index, ranking 49th of 744 measured models. The closest models on this index are GPT-5.6 Terra (42.1), Claude Opus 4.8 (batch) (41.8), and GLM-5.3-Flash (41.8). At $8 per million tokens (3:1 input/output blended price), it is more expensive than GPT-5.6 Terra ($4.50) and GLM-5.3-Flash ($0.237). With an output speed of 81 t/s, it is faster than GLM-5.3-Flash (61.1 t/s), slower than GPT-5.6 Terra (107 t/s).

Artificial Analysis Indices
High
Intelligence42.3
Rank 49 of 744 measured models
Coding77.2
Rank 7 of 298 measured models
Agentic44.2
Rank 33 of 189 measured models
Reasoning levels
Each point is a reasoning level of this model: more reasoning often raises the score but costs more. Select a point to open that level.
Intelligence Indexmost efficient ↗Cost per task
Capability indexes
Performance in specific industries, measured by Artificial Analysis
Finance & Accounting45.8
Rank 47 of 218 measured models
Strategy & Ops50.6
Rank 40 of 218 measured models
Legal47.3
Rank 52 of 218 measured models
Healthcare & Medical41.8
Rank 25 of 82 measured models
Engineering43.0
Rank 54 of 217 measured models
Economics49.8
Rank 59 of 225 measured models
Standard Benchmarks
GPQA Diamond92.8%
Rank 28 of 619 measured models
HLE46.0%
Rank 45 of 664 measured models
SciCode57.8%
Rank 23 of 231 measured models
IFBench69.2%
Rank 75 of 450 measured models
AA-LCR81.7%
Rank 50 of 575 measured models
TerminalBench Hard62.1%
Rank 5 of 432 measured models
Terminal-Bench 2.187.3%
Rank 20 of 244 measured models
Terminal-Bench v420.7%
Rank 59 of 218 measured models
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.

Preferences and analytics

Rybbit: visits, performance, clicks, copies, forms and session replay (masked inputs), with your permission. Refusing does not limit the site.

Privacy

Allow navigation analytics?

With your permission, Rybbit measures visits, performance and interactions (clicks, copies, forms), including session replay, to improve BenchSift. Input values are masked in replays. Refusing does not limit the site.

Your choice is stored for 180 days. Change it at any time using “Privacy preferences” in the footer. Withdrawal reloads the page.

No choice stored: Rybbit is disabled.

Privacy