Request a pilot
‹ PublicAI Index

The LLM benchmark aggregator.

Grok 3 Mini

SpaceXAI

Strongest in Safe-prompt compliance (#8 of 82), weakest in Harm refusal (#157 of 299). Above par in 19 of 21 scopes. Among the models it meets almost everywhere, it finishes behind Claude Opus 5 and GPT-5.5 and ahead of Grok 4.1 Fast and Kimi K2 Thinking.

§ 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.

Professional53.1#49/168
Legal53.8#47/151
Medical53.5#55/140
Finance53.3#66/151
Coding52.9#50/164
Code generation55.6#25/77
Agentic coding51.2#63/157
Core abilities52.9#62/204
General intelligence54.2#65/204
Knowledge50.2#84/138
Academic knowledge50.2#77/123
Reasoning50.6#87/178
Mathematics55#51/140
Science52.2#70/122
Reasoning42.8#99/140
Safety52.7#109/336
Safe-prompt compliance59.5#8/82
Fairness56.9#51/299
Harm refusal49.9#157/299
Human preference53.5#154/341
Human preference53.5#154/341

§ 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 Mini, left for the other.

Command R Pluswon 14 · lost 0
Mistral Smallwon 15 · lost 0
Llama 4 Scout Instructwon 18 · lost 0
Gemini 2.0 Flashwon 19 · lost 1
Command Awon 17 · lost 1
Gemini 1.5 Flashwon 13 · lost 1
Mistral Largewon 12 · lost 1
GPT-4o Miniwon 18 · lost 2
GPT-4.1 Nanowon 18 · lost 2
Mistral Large 3won 15 · lost 2
Gemini 1.5 Prowon 14 · lost 2
Mistral Medium 3.5won 13 · lost 2
DeepSeek V3won 18 · lost 3
GPT-4 Turbowon 11 · lost 2
GPT-4owon 16 · lost 3
GPT OSS 20Bwon 14 · lost 3
Claude 3.5 Haikuwon 12 · lost 3
GPT-4.1 Miniwon 16 · lost 4
Gemini 2.5 Flash Litewon 13 · lost 4
GLM-4.5won 12 · lost 4
Gemma 4 31Bwon 12 · lost 4
GPT OSS 120Bwon 15 · lost 5
Kimi K2won 15 · lost 5
Grok 4.3won 15 · lost 5
GPT-5.4 Nanowon 13 · lost 5
Mimo V2.5 Prowon 13 · lost 5
Qwen3.6 Pluswon 12 · lost 5
Grok 3won 14 · lost 6
Claude Haiku 4.5won 14 · lost 6
Ling 3.0 Flashwon 9 · lost 4
GPT-5 Nanowon 12 · lost 6
Gemini 2.5 Flashwon 14 · lost 7
Kimi K2.6won 13 · lost 7
GPT-5.4 Miniwon 11 · lost 6
Gemini 3.5 Flash Litewon 11 · lost 6
Nemotron 3 Ultrawon 9 · lost 5
MiniMax M2.1won 9 · lost 5
O3 Miniwon 12 · lost 7
Mimo V2.5won 8 · lost 5
GLM-4.6won 11 · lost 7
Qwen3 235B A22B Instructwon 12 · lost 8
MiniMax M3won 12 · lost 8
Qwen3.7 Maxwon 12 · lost 8
DeepSeek V3.2won 11 · lost 8
MiniMax M2.5won 9 · lost 7
Claude 3.5 Sonnetwon 10 · lost 8
GPT-4.1won 11 · lost 9
DeepSeek V4 Flashwon 11 · lost 9
Gemini 3.6 Flashwon 11 · lost 9
Kimi K2 Thinkingwon 9 · lost 8
Grok 4.1 Fastwon 9 · lost 8
Inkling Smallwon 7 · lost 7
MiniMax M2.7won 8 · lost 8
Gemini 3.1 Flash Litewon 8 · lost 8
Inklingwon 10 · lost 10
GLM-5.2won 10 · lost 10
Gemini 3.5 Flashwon 10 · lost 10
O4 Miniwon 10 · lost 11
Claude Sonnet 4won 10 · lost 11
Grok 4 Fastwon 8 · lost 9
Qwen3.8 27Bwon 8 · lost 9
GPT-6 Lunawon 8 · lost 9
Claude Opus 4.1won 6 · lost 7
GLM-5.3 Flashwon 9 · lost 11
Qwen3.5 Flashwon 6 · lost 8
GLM-4.7won 6 · lost 8
Kimi K2.5won 8 · lost 11
GLM-5.1won 7 · lost 10
Qwen3 Maxwon 6 · lost 9
Grok 4won 8 · lost 12
DeepSeek R1won 8 · lost 12
Claude Sonnet 4.6won 8 · lost 12
Gemini 3.1 Prowon 7 · lost 11
Claude Opus 4won 8 · lost 13
Grok 4.20won 6 · lost 10
GLM-5won 6 · lost 11
GPT-5.6 Lunawon 7 · lost 13
Gemini 3 Flashwon 5 · lost 10
Claude 3.7 Sonnetwon 6 · lost 12
O1won 6 · lost 12
DeepSeek V4.1 Flashwon 5 · lost 11
Gemini 3.8 Flashwon 6 · lost 14
Claude Opus 4.6won 6 · lost 14
GPT-6 Solwon 5 · lost 12
Claude Opus 4.5won 5 · lost 13
Claude Sonnet 4.5won 5 · lost 15
Gemini 3 Prowon 5 · lost 15
Mimo V2.6 Prowon 3 · lost 11
GPT-5.2won 4 · lost 16
DeepSeek V4 Prowon 4 · lost 16
GPT-5.4won 4 · lost 16
Qwen3.8 Maxwon 4 · lost 16
Muse Spark 1.1won 4 · lost 16
Gemini 2.5 Prowon 4 · lost 17
Grok 4.7won 3 · lost 14
GPT-5.1won 3 · lost 15
Muse Sparkwon 2 · lost 11
GPT-5 Miniwon 3 · lost 17
Grok 4.5won 3 · lost 17
GPT-5.6 Terrawon 3 · lost 17
GPT-6 Astrawon 2 · lost 13
Muse Spark 1.2won 2 · lost 14
GLM-5.3won 2 · lost 15
Gemini 3.7 Flashwon 2 · lost 15
GPT-5won 2 · lost 18
Grok 4.6won 2 · lost 18
Claude Sonnet 5won 2 · lost 18
O3won 2 · lost 19
Kimi K3won 1 · lost 16
GPT-5.6 Solwon 1 · lost 19
Claude Opus 4.7won 1 · lost 19
Muse Spark 1.3won 0 · lost 13
Claude Fable 5.1won 0 · lost 17
Claude Fable 5won 0 · lost 17
Claude Opus 5.5won 0 · lost 17
Claude Opus 4.8won 0 · lost 20
GPT-5.5won 0 · lost 20
Claude Opus 5won 0 · lost 20

§ 3 · Sources

Where the numbers come from

7 publications, 17 figures. Every one links to the page it was read from.

LMArena Text ↗1 measureread 2026-09-27
LMArena Text 1364
ARC-AGI-2 ↗1 measureread 2026-09-27
ARC-AGI-2 0.4%
LiveCodeBench ↗1 measureread 2026-09-27
LiveCodeBench 66.7%
Aider polyglot ↗1 measureread 2026-09-27
Aider polyglot 49.3%
Kagi LLM Benchmark ↗1 measureread 2026-09-27
Kagi LLM Benchmark 61.3%
Vals.ai ↗7 measuresread 2026-09-27
Vals · LegalBench 83.14%Vals · CorpFin 61.11%Vals · TaxEval 72.98%Vals · MedQA 90.1%Vals · GPQA Diamond 79.29%Vals · MMLU Pro 81.37%Vals · AIME 85%
HELM Safety ↗5 measuresread 2026-09-27
HELM Safety · HarmBench 57.2%HELM Safety · SimpleSafetyTests 99.3%HELM Safety · Anthropic Red Team 99.5%HELM Safety · BBQ 96.7%HELM Safety · XSTest 98.4%

Badge

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