Request a pilot
‹ PublicAI Index

The LLM benchmark aggregator.

Grok 3

SpaceXAI

Strongest in Factual grounding (#22 of 101), weakest in Safety (#297 of 336). Above par in 13 of 24 scopes. Among the models it meets almost everywhere, it finishes behind O3 and Claude Opus 5 and ahead of Qwen3.5 Flash and GPT-4 Turbo.

§ 1 · Profile

What it is good at

Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.

Professional52.5#54/168
Legal53.4#51/151
Finance54#60/151
Medical50.2#80/140
Coding51.9#59/164
Agentic coding52.3#57/157
Core abilities52.9#61/204
General intelligence54.2#64/204
Knowledge48.7#92/138
Academic knowledge48.4#85/123
Human preference58.1#113/341
Human preference58.1#113/341
Reasoning45.3#127/178
Science48.7#82/122
Mathematics48#87/140
Reasoning41.3#108/140
Safety44.1#297/336
Factual grounding60.5#22/101
Safe-prompt compliance55.1#36/82
Fairness51.4#97/299
Secure code40.9#212/274
Harm refusal43.5#265/299
Toxicity avoidance26#267/272
Jailbreak resistance26#272/272

§ 2 · Head to head

What it beats, and what beats it

The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for Grok 3, left for the other.

Mistral Medium 3.5won 13 · lost 2
Mistral Large 3won 15 · lost 3
Gemini 2.0 Flashwon 18 · lost 5
Command R Pluswon 14 · lost 4
GPT-4o Miniwon 17 · lost 5
Llama 4 Scout Instructwon 17 · lost 5
Command Awon 17 · lost 5
GPT-4.1 Nanowon 17 · lost 6
Gemini 1.5 Prowon 14 · lost 5
Mistral Smallwon 14 · lost 5
Jamba 1.6 Miniwon 11 · lost 4
Claude 3.5 Haikuwon 13 · lost 5
DeepSeek V3won 17 · lost 7
Gemini 1.5 Flashwon 12 · lost 5
Jamba 1.6 Largewon 10 · lost 5
Gemini 2.5 Flash Litewon 12 · lost 6
GPT-5.4 Nanowon 12 · lost 6
GPT-4.1 Miniwon 15 · lost 8
Gemini 3.5 Flash Litewon 10 · lost 6
GPT OSS 20Bwon 12 · lost 8
Claude 3.5 Sonnetwon 12 · lost 8
GPT-4owon 13 · lost 9
Mistral Largewon 10 · lost 7
GPT-5.4 Miniwon 10 · lost 7
GPT OSS 120Bwon 14 · lost 10
MiniMax M2.5won 9 · lost 7
Qwen3.6 Pluswon 9 · lost 7
Grok 4 Fastwon 10 · lost 8
Grok 4.3won 12 · lost 10
Claude Haiku 4.5won 13 · lost 11
GPT-4 Turbowon 8 · lost 7
Qwen3.5 Flashwon 8 · lost 7
O3 Miniwon 10 · lost 11
O1 Miniwon 7 · lost 8
Kimi K2won 11 · lost 13
GPT-5 Nanowon 10 · lost 12
Ling 3.0 Flashwon 7 · lost 9
Grok 4.1 Fastwon 9 · lost 12
Mimo V2.5 Prowon 9 · lost 12
GLM-4.5won 8 · lost 11
Gemma 4 31Bwon 8 · lost 11
Gemini 3.1 Flash Litewon 7 · lost 10
GLM-4.6won 9 · lost 13
MiniMax M3won 9 · lost 13
DeepSeek V3.2won 9 · lost 14
Kimi K2.6won 9 · lost 14
Qwen3.8 27Bwon 6 · lost 10
Grok 4.20won 6 · lost 10
GPT-4.1won 9 · lost 15
Qwen3.7 Maxwon 8 · lost 14
GLM-5.1won 7 · lost 13
Qwen3 Maxwon 5 · lost 10
GLM-5won 6 · lost 12
Gemini 2.5 Flashwon 8 · lost 16
Gemini 3.5 Flashwon 7 · lost 15
Gemini 3.6 Flashwon 7 · lost 15
MiniMax M2.7won 6 · lost 13
Gemini 3 Flashwon 5 · lost 11
Grok 4won 7 · lost 16
Kimi K2.5won 7 · lost 16
Grok 3 Miniwon 6 · lost 14
O4 Miniwon 7 · lost 17
Claude Sonnet 4won 7 · lost 17
Claude 3.7 Sonnetwon 6 · lost 15
Claude Sonnet 4.6won 6 · lost 15
Gemini 3.1 Prowon 5 · lost 13
DeepSeek V4 Flashwon 6 · lost 16
Inklingwon 6 · lost 16
GPT-6 Lunawon 5 · lost 14
DeepSeek R1won 6 · lost 17
Qwen3 235B A22B Instructwon 6 · lost 17
O1won 5 · lost 16
Claude Opus 4.6won 5 · lost 16
Claude Opus 4.5won 4 · lost 14
GPT-5 Miniwon 5 · lost 19
Claude Opus 4won 5 · lost 19
Claude Sonnet 4.5won 5 · lost 19
GPT-6 Solwon 4 · lost 16
Gemini 3 Prowon 4 · lost 17
GPT-5.6 Lunawon 4 · lost 18
GLM-5.3 Flashwon 4 · lost 18
GLM-5.2won 4 · lost 18
GPT-5.2won 4 · lost 19
Kimi K2 Thinkingwon 3 · lost 17
GPT-5.1won 3 · lost 19
Muse Spark 1.1won 3 · lost 19
GPT-5.4won 3 · lost 20
Gemini 2.5 Prowon 3 · lost 21
Grok 4.5won 2 · lost 20
DeepSeek V4 Prowon 2 · lost 21
Claude Opus 4.7won 2 · lost 21
Mimo V2.6 Flashwon 1 · lost 14
DeepSeek V4.1 Flashwon 1 · lost 14
GPT-6 Astrawon 1 · lost 14
Mimo V2.6 Prowon 1 · lost 16
Grok 4.7won 1 · lost 18
Claude Sonnet 5won 1 · lost 21
GPT-5.6 Terrawon 1 · lost 21
Qwen3.8 Maxwon 1 · lost 21
Gemini 3.8 Flashwon 1 · lost 21
GPT-5.5won 1 · lost 22
GPT-5.6 Solwon 1 · lost 22
GPT-5won 1 · lost 23
Muse Spark 1.2won 0 · lost 15
GLM-5.3won 0 · lost 16
Gemini 3.7 Flashwon 0 · lost 16
Kimi K3won 0 · lost 16
Claude Fable 5.1won 0 · lost 16
Claude Fable 5won 0 · lost 16
Claude Opus 5.5won 0 · lost 19
Grok 4.6won 0 · lost 22
Claude Opus 4.8won 0 · lost 22
Claude Opus 5won 0 · lost 22
O3won 0 · lost 23

§ 3 · Sources

Where the numbers come from

9 publications, 24 figures. Every one links to the page it was read from.

LMArena Text ↗1 measureread 2026-09-27
LMArena Text 1411
ARC-AGI-2 ↗1 measureread 2026-09-27
ARC-AGI-2 0%
Aider polyglot ↗1 measureread 2026-09-27
Aider polyglot 53.3%
Kagi LLM Benchmark ↗1 measureread 2026-09-27
Kagi LLM Benchmark 61.3%
SimpleBench ↗1 measureread 2026-09-27
SimpleBench 36.1%
Vals.ai ↗7 measuresread 2026-09-27
Vals · LegalBench 82.59%Vals · CorpFin 59.71%Vals · TaxEval 75.88%Vals · MedQA 83.85%Vals · GPQA Diamond 74.24%Vals · MMLU Pro 79.95%Vals · AIME 58.75%
HELM Safety ↗5 measuresread 2026-09-27
HELM Safety · HarmBench 45.3%HELM Safety · SimpleSafetyTests 96.8%HELM Safety · Anthropic Red Team 95.5%HELM Safety · BBQ 93.6%HELM Safety · XSTest 96.4%
Enkrypt AI Safety Leaderboard ↗6 measuresread 2026-09-27
Enkrypt · Jailbreak risk 68.2%Enkrypt · Harmful content risk 59.4%Enkrypt · CBRN risk 12%Enkrypt · Toxicity risk 38.3%Enkrypt · Bias risk 80.6%Enkrypt · Insecure code risk 52%
Vectara · Factual consistency 94.2%

Badge

PublicAI Index badge for Grok 3[![PublicAI Index](https://publicai.io/model-index/badge?model=grok-3)](https://publicai.io/model-index/m/grok-3)