readable AI benchmarks (simplified)

readable AI benchmarks

SpaceXAI

Grok 4.7 (Low)

reasoning · closed weights · Sep 21, 2026

Grok 4.7 (Low) has a Quality Score of 47.4, ranking 80th among 283 scored models. Per 1M tokens, pricing is $2.00 input, $6.00 output, and $0.500 cached input. It has a 500k token context window. Its Reliability Score is 63.7, ranking 62nd among 288 scored models, above average (total average: 48.2). Its Value Score is 32.0, around average compared with the total Value average of 29.3.

Quality Score47.4coverage 100.0%: AA Intelligence and AA-Omniscience outcomes available
Value Score32.0
Factual reliability63.7
Cache Discount75%

Model specification

Reasoning
reasoning
Input modalities
none
Output modalities
none
Context window
500k tokens
Weights
closed weights

Token prices USD per 1M tokens

Input
$2.00
Output
$6.00
Cached Input
$0.500
Cache write
not available
Cost per task
$2.92

Capability

Intelligence
31.1
Omniscience (native net)
27.4
Correct
45.7%
Blended price
$1.35

Answer outcomes

Fully graded outcomes (Correct + Incorrect): 64.1%

Correct 45.7%Incorrect 18.3%Partial / not attempted 35.9%

Artificial Analysis benchmarks aa-v5

Current Intelligence Index components

AA-Briefcase v1.1
1492.5 Elo
AutomationBench-AA
56.7%
private 657-task split from dataset v1.0.6
GDP.pdf
21.4%
FrontierCode v2.0
21.8%
AA Stirrup · 132 tasks, 3 attempts per task
Terminal-Bench v4.0
16.2%
AA mini-swe-agent standardized runs · 66 tasks, pass@1 averaged over 3 repeats per task
AA-Omniscience v1.1
27.4 native net
HLE-Diamond
18.5%
CritPt v2 Hard
43.5%
AA Stirrup · hard subset, 40 of 70 problems
Terminal-Bench-Science v0.1.0
4.8%

Standalone evaluations

GDPval-AA v2.1
1587.4 Elo

Historical evaluations

SciCode v1.0.1
54.9%
AA Long Context Reasoning v1.1
80.0%
Humanity's Last Exam
39.2%
CritPt
13.4%
FrontierCode
39.6%

Similar models

Support me! Patreon