ModelsCompareBest forBenchmarksStatusPricingAPI

Grok 4.6

xAI·Pending·Released 2026-08-12textvision
Context500k
Max output-
Input / 1M$2 ↗
Output / 1M$6 ↗
Cached in$0.5
Benchmarks15

Benchmarks

Cited public sources · rank is exact-benchmark, apples-to-apples

FrontierMath Tier 4 v2 (Epoch AI run)#16/4431.7% ↗
FrontierMath Tiers 1-3 v2 (Epoch AI run)#20/5966% ↗
GPQA Diamond (Epoch AI run)#7/7794% ↗
OTIS Mock AIME 2024-2025 (Epoch AI run)#6/7699.2% ↗
SimpleQA Verified (Epoch AI run)#17/5349.3% ↗
AA-Briefcase#1/21577 ↗
APEX-Agents#1/357.5% ↗
APEX-SWE56.4% ↗
Artificial Analysis Intelligence Index#1/361 ↗
CursorBench 3.269.9% ↗
DeepSWE v1.1#8/1365.9% ↗
FrontierCode v1.1 (Extended)61.3% ↗
GDPVal-AA v2#1/41753 ↗
Harvey LAB (Vals)15.8% ↗
Terminal-Bench 3.0#2/326% ↗

Compare

vs GPT-6 Astravs GPT-5.6 Solvs GPT-5.6 Terravs GPT-5.6 Lunavs Claude Fable 5.1vs Claude Opus 5vs Claude Sonnet 5vs Claude Haiku 4.5

More from xAI

Grok 4.3Grok 4.5Grok 4 Fast ReasoningGrok 4.20 ReasoningGrok 4.20 Non-ReasoningGrok 4.20 Multi-Agent

Access Grok 4.6 and every other model through one endpoint with automatic failover: Respan gateway.