Compare Models
Grok 4.6

Release Date: Aug 12, 2026

Developer SpaceXAI 🇺🇸
Context Window 500k
Max Output Tokens 500k
Token Costs (in/out) $2.00/6.00
Weights Private
Input Modalities

Accuracy

59.17 % ± 1.21

Cost / Test (Vals Index)

$ 4.335

Latency

38 min 4 s

Vals Index
BenchmarksAccuracyRankings
Vals Index

0.0%

±1.21
14/57
Code Migration

0.0%

±4.48
11/59
EMB

0.0%

±3.08
18/57
Finance Agent (v2)

0.0%

±0.67
20/60
Time Horizon Index: KSP

0.0%

±0.00
10/11
Legal Research Bench

0.0%

±3.47
6/60
MedCode

0.0%

±2.26
31/92
MedScribe

0.0%

±1.96
15/94
MortgageTax

0.0%

±0.95
47/98
ProofBench v1.1

0.0%

±5.02
14/31
SAGE

0.0%

±3.08
74/81
Public Benefits Bench v1.1

0.0%

±1.23
9/35
Tax Agent Bench

0.0%

±3.00
5/20
TaxEval v2

0.0%

±0.90
87/145
Vibe Code Bench v1.1

0.0%

±3.82
18/95
BioMysteryBench

0.0%

±0.00
3/11
Harvey's Legal Agent Benchmark

0.0%

±2.65
5/61
GPQA Diamond

0.0%

±1.13
3/138
IOI

0.0%

±2.09
16/27
LiveCodeBench

0.0%

±0.94
7/143
LegalBench

0.0%

±0.42
12/144
MMLU Pro

0.0%

±0.30
11/138
SkillsBench

0.0%

±4.84
13/35
SWE-bench

0.0%

±0.92
4/88
Terminal-Bench 2.1

0.0%

±2.08
10/65
Terminal-Bench Science

0.0%

±2.44
14/25
Proprietary BenchmarksAcademic BenchmarksIndustry Partners
Vals
Default Provider : SpaceXAI
Temperature: 0.7
Top P: 0.95
Top K: Default
Max Output Tokens: 500,000
Reasoning Effort: high