The LLM benchmark aggregator.

Llama 3.1 8B Instruct

Meta · 8B · open weights

Strongest in Toxicity avoidance (#1 of 266), weakest in Agents (#293 of 298). Above par in 3 of 15 scopes. Among the models it meets almost everywhere, it finishes behind Gemini 3 Pro and Claude Sonnet 4.6 and ahead of Jamba 1.5 Mini and Mixtral 8x22B Instruct.

§ 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.

Safety49.7−11.7#185/3272/3

Toxicity avoidance59leads#1/2661/1

Safe-prompt compliance52.6−7.8#48/821/1

Jailbreak resistance56.1−8.9#99/2661/1

Harm refusal49.4−12.1#157/2902/2

Secure code45.6−20.6#189/2681/1

Fairness43−27#234/2902/2

Professional38.2−24.2#185/1941/2

Medical38.9−27.6#134/1491/1

Finance30−32.4#161/1621/1

Human preference38.5−30.3#284/3541/1

Agents39.8−26.5#293/2982/6

Tool use41.3−32.7#62/811/1

Knowledge work30−40.7#203/2081/2

§ 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 Llama 3.1 8B Instruct, left for the other.

DBRX Instructwon 8 · lost 1

GPT-3.5 Turbowon 8 · lost 1

Mixtral 8x7B Instructwon 9 · lost 2

Llama 3.2 3B Instructwon 8 · lost 2

Jamba 1.6 Miniwon 7 · lost 2

DeepSeek LLM 67B Chatwon 7 · lost 2

Jamba 1.6 Largewon 6 · lost 3

Command R+won 7 · lost 4

Llama 3.2 1B Instructwon 6 · lost 4

Gemma 3 4B ITwon 6 · lost 4

Qwen3 8Bwon 5 · lost 4

Mixtral 8x22B Instructwon 6 · lost 5

Jamba 1.5 Miniwon 6 · lost 5

Qwen3 30B A3B Instructwon 5 · lost 5

Llama 3.1 70B Instructwon 6 · lost 6

Magistral Small 1.2won 5 · lost 6

Jamba 1.5 Largewon 5 · lost 6

Gemini 1.5 Flashwon 5 · lost 6

Llama 2 13B Chatwon 4 · lost 5

Gemma 2 9B ITwon 4 · lost 5

Gemini 2.0 Flash-Litewon 4 · lost 5

Grok 3won 5 · lost 7

Gemma 4 31Bwon 5 · lost 7

Granite 3.1 8B Instructwon 4 · lost 6

Amazon Nova Prowon 4 · lost 6

NVIDIA Nemotron 3 Super 120B A12Bwon 4 · lost 6

DeepSeek V3.1won 4 · lost 6

Llama 4 Scout Instructwon 6 · lost 9

Command Awon 5 · lost 8

MiMo V2.6 Flashwon 5 · lost 8

Ling 3.0 Flashwon 4 · lost 7

Gemma 3 12B ITwon 4 · lost 7

GLM-4.5won 4 · lost 7

DeepSeek V3won 5 · lost 9

DeepSeek R1won 5 · lost 9

Kimi K2 Thinkingwon 3 · lost 6

Llama 2 7B Chatwon 3 · lost 6

Llama 2 70B Chatwon 3 · lost 6

Claude 3 Haikuwon 3 · lost 6

Llama 3.3 70B Instructwon 4 · lost 9

Mistral Large 4won 4 · lost 9

NVIDIA Nemotron 3 Ultra 550B A55Bwon 4 · lost 9

Grok 4.1 Fastwon 4 · lost 9

MiniMax M3won 4 · lost 9

DeepSeek V4 Flashwon 4 · lost 9

Grok 4.3won 4 · lost 9

Kimi K2.5won 4 · lost 9

Kimi K2.6won 4 · lost 9

MiMo V2.5 Prowon 4 · lost 9

Qwen3.7 Maxwon 4 · lost 9

GLM-5.2won 4 · lost 9

Gemini 3.5 Flashwon 4 · lost 9

Gemini 3.6 Flashwon 4 · lost 9

Mercury 2won 3 · lost 7

Gemma 4 26B A4Bwon 3 · lost 7

Grok 4won 4 · lost 10

DeepSeek V3.2won 4 · lost 10

GLM-4.6won 4 · lost 10

Gemini 1.5 Prowon 3 · lost 9

GLM-5.3 Flashwon 3 · lost 10

Gemini 3.8 Flashwon 3 · lost 10

Llama 3 8B Instructwon 2 · lost 7

Claude 3 Sonnetwon 2 · lost 7

Claude 3 Opuswon 2 · lost 7

GPT OSS 20Bwon 3 · lost 11

GPT OSS 120Bwon 3 · lost 11

Gemini 2.5 Flashwon 3 · lost 11

Gemini 2.5 Prowon 3 · lost 11

Amazon Nova Microwon 2 · lost 8

Qwen3.5 35B A3Bwon 2 · lost 8

Qwen3.5 122B A10Bwon 2 · lost 8

Qwen3.5 397B A17Bwon 2 · lost 8

Llama 4 Maverick Instructwon 3 · lost 12

Claude Haiku 4.5won 3 · lost 12

Gemma 3 27B ITwon 2 · lost 9

Claude 3.5 Haikuwon 2 · lost 10

Llama 3.1 405B Instructwon 2 · lost 10

Claude 3.7 Sonnetwon 2 · lost 11

GPT-6 Lunawon 2 · lost 11

GPT-5.6 Lunawon 2 · lost 11

GPT-6 Solwon 2 · lost 11

DeepSeek V4 Prowon 2 · lost 11

GLM-5.1won 2 · lost 11

Grok 4.5won 2 · lost 11

GPT-5.6 Terrawon 2 · lost 11

GPT-5.5won 2 · lost 11

Qwen3.8 Maxwon 2 · lost 11

GPT-5.6 Solwon 2 · lost 11

Claude Opus 5won 2 · lost 11

O3 Miniwon 2 · lost 12

Kimi K2won 2 · lost 12

Qwen3 235B A22B Instructwon 2 · lost 12

GPT-4.1 nanowon 2 · lost 13

GPT-4.1 miniwon 2 · lost 13

Claude Sonnet 4.5won 2 · lost 13

Qwen2 72B Instructwon 1 · lost 8

Claude Opus 4.6won 1 · lost 8

Phi 4won 1 · lost 9

GPT-4won 1 · lost 9

Hy3won 1 · lost 9

Qwen2.5 72B Instructwon 1 · lost 10

GPT-4 Turbowon 1 · lost 10

O1 Miniwon 1 · lost 10

Gemini 2.0 Flashwon 1 · lost 10

Claude 3.5 Sonnetwon 1 · lost 11

O1won 1 · lost 11

Claude Opus 4won 1 · lost 11

MiniMax M2.7won 1 · lost 12

GPT-5.2won 1 · lost 12

Grok 4.7won 1 · lost 12

Inklingwon 1 · lost 12

Grok 4.6won 1 · lost 12

Claude Sonnet 5won 1 · lost 12

GPT-5.4won 1 · lost 12

MiMo V2.6 Prowon 1 · lost 12

Claude Opus 4.8won 1 · lost 12

Muse Spark 1.1won 1 · lost 12

Claude Opus 4.7won 1 · lost 12

Claude Opus 5.5won 1 · lost 12

GPT-4o miniwon 1 · lost 13

GPT-4owon 1 · lost 13

GPT-5 nanowon 1 · lost 13

O4 Miniwon 1 · lost 13

Claude Sonnet 4won 1 · lost 13

GPT-4.1won 1 · lost 13

O3won 1 · lost 13

GPT-5won 1 · lost 13

GPT-5.1won 1 · lost 13

GPT-5 miniwon 1 · lost 14

Grok 3 Miniwon 0 · lost 9

Gemini 2.5 Flash-Litewon 0 · lost 11

Claude Sonnet 4.6won 0 · lost 11

Gemini 3 Prowon 0 · lost 11

§ 3 · Sources

Where the numbers come from

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

LMArena Text ↗1 measureread 2026-11-11

LMArena Text 1211

GDPval-AA ↗1 measureread 2026-11-11

GDPval-AA -380

BFCL ↗1 measureread 2026-11-11

BFCL v4 25.83%

Vals.ai ↗3 measuresread 2026-11-11

Vals · CorpFin 37.8%Vals · TaxEval 32.34%Vals · MedQA 62.61%

HELM Safety ↗5 measuresread 2026-11-11

HELM Safety · HarmBench 61.6%HELM Safety · SimpleSafetyTests 98.8%HELM Safety · Anthropic Red Team 96.8%HELM Safety · BBQ 78.5%HELM Safety · XSTest 95.3%

Enkrypt AI Safety Leaderboard ↗6 measuresread 2026-11-11

Enkrypt · Jailbreak risk 7.7%Enkrypt · Harmful content risk 26.7%Enkrypt · CBRN risk 16%Enkrypt · Toxicity risk 0%Enkrypt · Bias risk 81.7%Enkrypt · Insecure code risk 42.2%

Badge

[![PublicAI Index](https://publicai.io/model-index/badge?model=llama-3-1-8b-instruct)](https://publicai.io/model-index/m/llama-3-1-8b-instruct)