Back

OpenAI

GPT-6 Sol

ReasoningClosed weights
Intelligence
44.2
Rank 41 of 744 measured models
Output speed
83.6 t/s
Rank 216 of 403 measured models
Price/1M
$4
Rank 532 of 659 measured models
Context window
1.1M
Rank 21 of 1199 measured models
Cost per task
$0.521
Rank 107 of 193 measured models

GPT-6 Sol (Xhigh) scores 44.2 on the Artificial Analysis Intelligence Index, ranking 41st of 744 measured models. The closest models on this index are GLM-5.3 (44.8), Grok 4.6 (44.3), and Step 5 Preview (43.7). At $4 per million tokens (3:1 input/output blended price), it is more expensive than GLM-5.3 ($2.15), Grok 4.6 ($3), and Step 5 Preview ($1.43). With an output speed of 83.6 t/s, it is faster than GLM-5.3 (61.1 t/s) and Grok 4.6 (50 t/s), slower than Step 5 Preview (89.3 t/s).

Artificial Analysis Indices
Extra high
Intelligence44.2
Rank 41 of 744 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 & Accounting46.5
Rank 42 of 218 measured models
Strategy & Ops50.9
Rank 38 of 218 measured models
Legal49.5
Rank 42 of 218 measured models
Engineering45.1
Rank 39 of 217 measured models
Economics52.5
Rank 44 of 225 measured models
Standard Benchmarks
HLE46.3%
Rank 43 of 664 measured models
SciCode55.1%
Rank 50 of 231 measured models
AA-LCR81.3%
Rank 54 of 575 measured models
Terminal-Bench v430.3%
Rank 44 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