Back

SpaceXSI

Grok 4.6

ReasoningClosed weights
Intelligence
35.1
Rank 90 of 744 measured models
Coding
66.3
Rank 67 of 298 measured models
Output speed
50.6 t/s
Rank 319 of 403 measured models
Price/1M
$3
Rank 487 of 659 measured models
Context window
500K
Rank 349 of 1199 measured models
Cost per task
$0.376
Rank 96 of 193 measured models

Grok 4.6 (Low) scores 35.1 on the Artificial Analysis Intelligence Index, ranking 90th of 744 measured models. The closest models on this index are Agnes 3.0 Flash (35.5), Agnes 2.5 Pro Beta (35.2), and DeepSeek V4 Flash Vision (Max) (34.8). At $3 per million tokens (3:1 input/output blended price), it is more expensive than Agnes 3.0 Flash ($0.075), Agnes 2.5 Pro Beta ($0.15), and DeepSeek V4 Flash Vision (Max) ($0.66). With an output speed of 50.6 t/s, it is slower than Agnes 2.5 Pro Beta (120 t/s) and DeepSeek V4 Flash Vision (Max) (173 t/s).

Artificial Analysis Indices
Low
Intelligence35.1
Rank 90 of 744 measured models
Coding66.3
Rank 67 of 298 measured models
Agentic41.2
Rank 42 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 & Accounting38.4
Rank 83 of 218 measured models
Strategy & Ops43.5
Rank 71 of 218 measured models
Legal43.8
Rank 64 of 218 measured models
Engineering35.0
Rank 88 of 217 measured models
Economics45.4
Rank 79 of 225 measured models
Standard Benchmarks
GPQA Diamond87.9%
Rank 97 of 619 measured models
HLE27.6%
Rank 176 of 664 measured models
SciCode49.4%
Rank 115 of 231 measured models
AA-LCR80.7%
Rank 64 of 575 measured models
Terminal-Bench 2.175.3%
Rank 63 of 244 measured models
Terminal-Bench v43.0%
Rank 105 of 218 measured models
τ-Bench Banking38.1%
Rank 40 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