The LLM benchmark aggregator.

Grok 4.20 (Reasoning, 0309)

SpaceXAI

Strongest in Core abilities (#19 of 225, on 1 of its 4 boards), weakest in Coding (#163 of 176). Above par in 10 of 16 scopes. Among the models it meets almost everywhere, it finishes behind Claude Fable 5 and Claude Fable 5.1 and ahead of MiMo V2.5 Pro and Gemma 4 31B.

§ 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.

Core abilities57.7−7.5#19/2251/4

General intelligence61.1−7.7#21/2141/3

Reasoning57.5−7.4#25/1902/5

Mathematics58−6.1#21/1501/2

Science58.8−4.7#34/1221/1

Reasoning55.5−11.2#38/1441/3

Human preference63.7−4.9#40/3501/1

Knowledge52.6−10.1#72/1911/3

Academic knowledge56.3−7.6#43/1231/1

Professional44−19.1#152/1881/2

Finance44.3−18.2#125/1611/1

Medical43.1−23.4#125/1491/1

Legal43.1−22.8#130/1641/1

Coding40.2−26.2#163/1761/6

Agentic coding39.6−27.4#157/1691/5

§ 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 4.20 (Reasoning, 0309), left for the other.

Laguna XS.2won 10 · lost 0

Mistral Smallwon 10 · lost 0

Command R+ (08-2024)won 10 · lost 0

Gemini 1.5 Flashwon 10 · lost 0

Mistral Small 3.1 24B Instructwon 11 · lost 0

NVIDIA-Nemotron-3.5-Lightning-30B-A3B (NVFP4)won 13 · lost 0

Mistral Medium 3.5won 15 · lost 0

Llama 4 Scout Instructwon 13 · lost 1

GPT OSS 20Bwon 12 · lost 1

Mercury 2.5won 9 · lost 1

Mistral Medium 3.1won 9 · lost 1

GPT-4.1 nanowon 14 · lost 2

GPT-4o miniwon 13 · lost 2

GPT-5 nanowon 12 · lost 2

Gemini 1.5 Prowon 10 · lost 2

Llama 3.3 70B Instructwon 9 · lost 2

Laguna M.1won 8 · lost 2

Command A (03-2025)won 12 · lost 3

Claude 3.5 Haikuwon 9 · lost 3

Llama 4 Maverick Instructwon 12 · lost 4

Grok 4 Fastwon 12 · lost 4

Mistral Large 3won 12 · lost 4

Claude Haiku 4.5won 12 · lost 4

NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)won 12 · lost 4

GPT-5.4 miniwon 11 · lost 4

Grok 2won 8 · lost 3

Grok 4.1 Fastwon 10 · lost 4

Mistral Largewon 7 · lost 3

Gemini 2.5 Flash-Litewon 9 · lost 4

GLM-4.5won 9 · lost 4

MiMo V2.5won 9 · lost 4

GPT OSS 120Bwon 11 · lost 5

Gemini 2.0 Flashwon 11 · lost 5

GPT-5.4 nanowon 11 · lost 5

DeepSeek R1won 10 · lost 5

Qwen3 235B A22B Instructwon 10 · lost 5

Gemini 3.1 Flash-Litewon 10 · lost 5

O3 Miniwon 9 · lost 5

GLM-4.7won 9 · lost 5

Grok 4.3won 9 · lost 5

Grok 3 Miniwon 10 · lost 6

GPT-4.1 miniwon 10 · lost 6

GPT-5 miniwon 10 · lost 6

MiniMax M2.5won 10 · lost 6

O4 Miniwon 10 · lost 6

DeepSeek V3won 10 · lost 6

Claude Sonnet 4won 10 · lost 6

Inkling Smallwon 10 · lost 6

Gemini 2.5 Flashwon 10 · lost 6

Grok 3won 10 · lost 6

Grok 4won 10 · lost 6

GPT-4.1won 10 · lost 6

Kimi K2won 10 · lost 6

DeepSeek V3.2won 10 · lost 6

Claude Opus 4won 10 · lost 6

O3won 10 · lost 6

Gemini 2.5 Prowon 10 · lost 6

Gemini 3.5 Flash-Litewon 10 · lost 6

Qwen3.5 Flashwon 8 · lost 5

MiniMax M2.7won 8 · lost 5

Qwen3 Maxwon 9 · lost 6

GLM-4.6won 9 · lost 6

Step 5won 7 · lost 5

Kimi K2 Thinkingwon 8 · lost 6

Claude 3.7 Sonnetwon 8 · lost 6

O1won 8 · lost 6

MiniMax M3won 9 · lost 7

Claude 3.5 Sonnetwon 7 · lost 6

Mistral Large 4won 7 · lost 6

Gemma 4 31Bwon 7 · lost 6

MiMo V2.5 Prowon 7 · lost 6

MiMo V2.6 Flashwon 6 · lost 6

Claude Opus 4.1won 6 · lost 6

GPT-4owon 7 · lost 7

GPT-6 Lunawon 7 · lost 7

GPT-5won 8 · lost 8

Qwen3.8 27Bwon 8 · lost 8

GPT-5.2won 8 · lost 8

DeepSeek V4 Flashwon 8 · lost 8

Inklingwon 8 · lost 8

Kimi K2.5won 8 · lost 8

GPT-5.6 Lunawon 8 · lost 8

Claude Sonnet 4.5 (High, 32K, 20250929)won 8 · lost 8

GLM-5won 8 · lost 8

Grok 4.7won 6 · lost 7

Claude Sonnet 5won 7 · lost 9

DeepSeek-V4-Pro (High, 20260813)won 7 · lost 9

GLM-5.3 Flashwon 7 · lost 9

MiniMax M2.1won 6 · lost 8

Qwen3.6 Pluswon 6 · lost 8

GPT-5.1won 6 · lost 8

GLM-5.1won 6 · lost 8

Claude Haiku 5.5won 5 · lost 7

Ling 3.0 Flashwon 4 · lost 6

MiMo V2.6 Prowon 5 · lost 8

Claude Sonnet 4.6won 6 · lost 10

GLM-5.2won 6 · lost 10

GLM-5.3won 6 · lost 10

Kimi K2.6won 5 · lost 9

Grok 4.6won 5 · lost 11

Grok 4.5won 5 · lost 11

GPT-5.6 Terrawon 5 · lost 11

Claude Opus 4.5 (High, 32K, 20251101)won 5 · lost 11

Qwen3.8 Maxwon 5 · lost 11

GPT-6 Solwon 4 · lost 10

Gemini 3 Flashwon 4 · lost 10

Qwen 3.5 Pluswon 3 · lost 8

Gemini 3.6 Flashwon 4 · lost 11

Muse Spark 1.2won 4 · lost 11

Gemini 3.1 Prowon 4 · lost 12

Kimi K3won 4 · lost 12

Gemini 3.7 Flashwon 4 · lost 12

Claude Opus 4.6won 4 · lost 12

Muse Spark 1.1won 3 · lost 11

Qwen3.7 Maxwon 3 · lost 12

DeepSeek V4.1 Flashwon 3 · lost 12

Gemini 3.8 Flashwon 3 · lost 13

DeepSeek V4won 2 · lost 9

Muse Spark 1.3won 2 · lost 10

Gemini 3.5 Flashwon 2 · lost 13

GPT-5.4won 2 · lost 14

Claude Opus 4.8won 2 · lost 14

Claude Opus 5won 2 · lost 14

Claude Opus 4.7won 2 · lost 14

Muse Wonder Fullwon 1 · lost 9

Muse Sparkwon 1 · lost 12

Claude Sonnet 5.5won 1 · lost 13

GPT-5.6 Solwon 1 · lost 15

Muse Wonderwon 0 · lost 12

Gemini 4 Argonwon 0 · lost 13

GPT-6 Astrawon 0 · lost 14

GPT-6.1 Solwon 0 · lost 14

Claude Opus 5.5won 0 · lost 14

GPT-5.5won 0 · lost 16

Gemini 3 Prowon 0 · lost 16

Claude Fable 5.1won 0 · lost 16

Claude Fable 5won 0 · lost 16

§ 3 · Sources

Where the numbers come from

4 publications, 19 figures. Every one links to the page it was read from.

LMArena Text ↗1 measureread 2026-10-21

LMArena Text 1472

ARC-AGI-2 ↗1 measureread 2026-10-21

ARC-AGI-2 65.1%

Kagi LLM Benchmark ↗1 measureread 2026-10-21

Kagi LLM Benchmark 75%

Vals.ai ↗16 measuresread 2026-10-21

Vals · Legal Research Bench 13.94%Vals · CaseLaw 54.45%Vals · LegalBench 77.74%Vals · Finance Agent 28.49%Vals · CorpFin 63.67%Vals · TaxEval 74.12%Vals · MortgageTax 45.35%Vals · MedQA 94.55%Vals · MedCode 32.16%Vals · MedScribe 63.41%Vals · SWE-bench Verified 72.2% (Mini-SWE-agent)Vals · Vibe Code Bench 4.06% (OpenHands)Vals · Code Migration 0.31%Vals · GPQA Diamond 88.64%Vals · MMLU Pro 86.25%Vals · AIME 96.46%

Badge

[![PublicAI Index](https://publicai.io/model-index/badge?model=grok-4-20-reasoning-0309)](https://publicai.io/model-index/m/grok-4-20-reasoning-0309)