SpaceXSI

Grok 4.6

Intelligence
31.0
Rank 119 of 788 measured models
Coding
75.9
Rank 22 of 301 measured models
Output speed
64.7 t/s
Rank 318 of 454 measured models
Price/1M
$3
Rank 550 of 741 measured models
Context window
500K
Rank 395 of 1333 measured models
Cost per task
$4.68
Rank 210 of 235 measured models

Grok 4.6 (Xhigh) scores 31.0 on the Artificial Analysis Intelligence Index, ranking 119th of 788 measured models. The closest models on this index are Muse Spark (31.3), Gemini 3.8 Flash (30.8), and Inkling 2 Small (30.6). At $3 per million tokens (3:1 input/output blended price), it is more expensive than Gemini 3.8 Flash ($1.50). With an output speed of 64.7 t/s, it is slower than Gemini 3.8 Flash (113 t/s) and Inkling 2 Small (88.7 t/s).

Artificial Analysis Indices
Extra high
Intelligence31.0
Rank 119 of 788 measured models
Coding75.9
Rank 22 of 301 measured models
Agentic52.5
Rank 10 of 191 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 & Accounting47.6
Rank 48 of 260 measured models
Strategy & Ops52.8
Rank 41 of 260 measured models
Legal43.5
Rank 72 of 260 measured models
Healthcare & Medical44.1
Rank 40 of 190 measured models
Engineering37.4
Rank 91 of 259 measured models
Economics42.0
Rank 92 of 267 measured models
Standard Benchmarks
GPQA Diamond93.5%
Rank 14 of 621 measured models
HLE44.1%
Rank 55 of 682 measured models
SciCode53.0%
Rank 87 of 249 measured models
AA-LCR81.0%
Rank 66 of 593 measured models
Terminal-Bench 2.188.0%
Rank 12 of 246 measured models
Terminal-Bench 4.017.2%
Rank 85 of 246 measured models
τ-Bench Banking43.3%
Rank 21 of 211 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