OpenAI

GPT-5.6 Sol

Intelligence
44.0
Rank 44 of 790 measured models
Coding
78.3
Rank 4 of 302 measured models
Output speed
97.2 t/s
Rank 189 of 456 measured models
Price/1M
$8
Rank 673 of 741 measured models
Context window
1M
Rank 107 of 1368 measured models
Cost per task
$1.18
Rank 153 of 235 measured models

GPT-5.6 Sol (Xhigh) scores 44.0 on the Artificial Analysis Intelligence Index, ranking 44th of 790 measured models. The closest models on this index are Claude Opus 5 (44.2), GPT-6 Sol (44.2), and Muse Wonder (43.6). At $8 per million tokens (3:1 input/output blended price), it is cheaper than Claude Opus 5 ($10), priced close to Muse Wonder ($8), more expensive than GPT-6 Sol ($4). With an output speed of 97.2 t/s, it is faster than Claude Opus 5 (81.2 t/s), slower than GPT-6 Sol (111 t/s) and Muse Wonder (110 t/s).

Artificial Analysis Indices
Extra high
Intelligence44.0
Rank 44 of 790 measured models
Coding78.3
Rank 4 of 302 measured models
Agentic47.4
Rank 25 of 192 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 & Accounting47.1
Rank 52 of 260 measured models
Strategy & Ops52.2
Rank 44 of 260 measured models
Legal48.9
Rank 42 of 260 measured models
Healthcare & Medical44.1
Rank 40 of 190 measured models
Engineering44.6
Rank 52 of 259 measured models
Economics51.4
Rank 40 of 267 measured models
Standard Benchmarks
GPQA Diamond93.1%
Rank 24 of 658 measured models
HLE47.3%
Rank 42 of 719 measured models
SciCode57.1%
Rank 44 of 286 measured models
IFBench71.0%
Rank 69 of 487 measured models
AA-LCR82.3%
Rank 44 of 630 measured models
TerminalBench Hard61.4%
Rank 6 of 432 measured models
Terminal-Bench 2.189.5%
Rank 5 of 246 measured models
Terminal-Bench 4.024.7%
Rank 71 of 246 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