Request a pilot
‹ PublicAI Index

The LLM benchmark aggregator.

Gemini 3.8 Flash

Google

Strongest in Instruction following (#1 of 57), weakest in Jailbreak resistance (#200 of 272). Above par in 24 of 30 scopes. Among the models it meets almost everywhere, it finishes behind Claude Fable 5.1 and Claude Fable 5 and ahead of DeepSeek V4.1 Flash and GPT-5.6 Terra.

§ 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.

Knowledge59.4#6/138
Academic knowledge61.2#5/123
Human preference65.9#8/342
Human preference65.9#8/342
Reasoning59.7#13/178
Science62.8#4/122
Reasoning65.5#6/140
Mathematics51.4#77/140
Core abilities56.1#32/204
Instruction following73.7#1/57
Language66#5/57
General intelligence57.3#41/204
Data analysis26#56/57
Agents58.9#38/268
Knowledge work61.6#32/178
Professional53.7#47/168
Cybersecurity29.9#8/9
Finance58.7#11/151
Biology research37.5#13/16
Legal58.8#18/151
Medical56.8#26/140
Coding51.5#63/165
Agentic coding54.4#43/157
Code generation40#61/77
Safety52.2#126/337
Secure code63#33/274
Fairness56.5#55/300
Toxicity avoidance54#129/272
Harm refusal47.8#193/300
Jailbreak resistance46.8#200/272

§ 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 Gemini 3.8 Flash, left for the other.

Mistral Large 3won 19 · lost 0
Grok 3won 20 · lost 0
DeepSeek V3won 25 · lost 0
DeepSeek V3.2won 23 · lost 1
Mimo V2.5 Prowon 22 · lost 1
Command Awon 21 · lost 1
Gemini 3.5 Flash Litewon 21 · lost 1
Grok Build 0.1won 18 · lost 1
Mistral Smallwon 18 · lost 1
GLM-4.5won 18 · lost 1
Ling 3.0 Flashwon 17 · lost 1
Gemini 3.1 Flash Litewon 17 · lost 1
Grok 4.3won 26 · lost 2
Kimi K2.6won 26 · lost 2
Qwen3.7 Maxwon 26 · lost 2
GLM-4.6won 21 · lost 2
Gemini 2.0 Flashwon 20 · lost 2
Qwen3.6 Pluswon 20 · lost 2
GPT-5.4 Miniwon 20 · lost 2
GPT OSS 20Bwon 19 · lost 2
Grok 4.1 Fastwon 19 · lost 2
Gemma 4 31Bwon 19 · lost 2
Gemini 3.6 Flashwon 26 · lost 3
Gemini 1.5 Prowon 16 · lost 2
GPT-4.1 Nanowon 21 · lost 3
GPT-4.1 Miniwon 21 · lost 3
Gemini 2.5 Flashwon 21 · lost 3
GPT-5.4 Nanowon 20 · lost 3
Kimi K2won 20 · lost 3
Llama 4 Scout Instructwon 19 · lost 3
Kimi K2.7 Codewon 16 · lost 3
GPT-4o Miniwon 20 · lost 4
Grok 4won 19 · lost 4
MiniMax M3won 23 · lost 5
GLM-5.2won 23 · lost 5
Gemini 3.5 Flashwon 23 · lost 5
GLM-5.1won 18 · lost 4
GPT-4owon 17 · lost 4
Claude Sonnet 4.6won 21 · lost 5
Gemini 2.5 Prowon 20 · lost 5
GPT OSS 120Bwon 19 · lost 5
Kimi K2.5won 19 · lost 5
Claude 3.5 Haikuwon 15 · lost 5
Claude 3.5 Sonnetwon 15 · lost 5
Claude Haiku 4.5won 18 · lost 6
DeepSeek R1won 18 · lost 6
Inklingwon 21 · lost 7
O3 Miniwon 17 · lost 6
GPT-4.1won 17 · lost 6
Qwen3 235B A22B Instructwon 17 · lost 6
DeepSeek V4 Flashwon 22 · lost 8
Qwen3.8 27Bwon 16 · lost 6
Claude Opus 4.5won 16 · lost 6
O4 Miniwon 17 · lost 7
Claude Sonnet 4.5won 17 · lost 7
Kimi K2 Thinkingwon 14 · lost 6
GPT-5.6 Lunawon 20 · lost 9
Gemini 3.7 Flashwon 15 · lost 7
GLM-5.3 Flashwon 19 · lost 9
GPT-5 Nanowon 14 · lost 7
Claude 3.7 Sonnetwon 14 · lost 7
MiniMax M2.7won 14 · lost 7
GPT-5 Miniwon 16 · lost 8
O3won 16 · lost 8
DeepSeek V4 Prowon 18 · lost 11
Gemini 3.1 Prowon 14 · lost 9
GPT-5.4won 17 · lost 11
O1won 12 · lost 8
Gemini 3 Prowon 12 · lost 8
Claude Sonnet 4won 15 · lost 10
GPT-5.2won 16 · lost 11
GPT-5.1won 13 · lost 9
Claude Opus 4.6won 14 · lost 10
Grok 4.5won 16 · lost 12
Muse Spark 1.1won 16 · lost 12
Claude Opus 4won 13 · lost 10
GLM-5.3won 13 · lost 10
GPT-6 Lunawon 14 · lost 12
GPT-5.6 Terrawon 15 · lost 13
DeepSeek V4.1 Flashwon 12 · lost 11
Claude Sonnet 5won 14 · lost 14
Qwen3.8 Maxwon 13 · lost 15
GPT-5.5won 13 · lost 15
Claude Opus 4.7won 13 · lost 15
GPT-6 Solwon 12 · lost 14
GPT-5won 11 · lost 13
Claude Opus 4.8won 12 · lost 16
Muse Spark 1.2won 9 · lost 13
GPT-5.6 Solwon 12 · lost 18
Kimi K3won 9 · lost 15
Mimo V2.6 Prowon 7 · lost 13
Grok 4.6won 10 · lost 20
GPT-6 Astrawon 7 · lost 15
Grok 4.7won 7 · lost 19
Claude Opus 5won 5 · lost 25
Muse Spark 1.3won 3 · lost 16
Claude Opus 5.5won 4 · lost 22
Claude Fable 5won 3 · lost 19
Claude Fable 5.1won 2 · lost 21

§ 3 · Sources

Where the numbers come from

10 publications, 37 figures. Every one links to the page it was read from.

LMArena Text ↗1 measureread 2026-09-26
LMArena Text 1492
Artificial Analysis ↗1 measureread 2026-09-26
Artificial Analysis Intelligence Index 41
GDPval-AA ↗1 measureread 2026-09-26
GDPval-AA 1412
AA-Briefcase ↗1 measureread 2026-09-26
AA-Briefcase 1202
Terminal-Bench ↗1 measureread 2026-09-26
Terminal-Bench 19.1% (mini-SWE-agent)
ARC-AGI-2 ↗1 measureread 2026-09-26
ARC-AGI-2 89.2%
LiveBench ↗8 measuresread 2026-09-26
LiveBench 75.8LiveBench · Reasoning 89.3LiveBench · Coding 72.5LiveBench · Agentic Coding 54.2LiveBench · Mathematics 91.6LiveBench · Data Analysis 54LiveBench · Language 87.8LiveBench · Instruction Following 81.4
SimpleBench ↗1 measureread 2026-09-26
SimpleBench 82.4%
Vals.ai ↗16 measuresread 2026-09-26
Vals · Legal Research Bench 38.94%Vals · LegalBench 86.99%Vals · Harvey Legal Agent Benchmark 10%Vals · Finance Agent 61.44%Vals · TaxEval 74.45%Vals · MortgageTax 65.34%Vals · MedCode 48.13%Vals · MedScribe 84.5%Vals · BioMysteryBench 62.22%Vals · CyberBench 43.75%Vals · SWE-bench Verified 80% (Mini-SWE-agent)Vals · Vibe Code Bench 78.65% (OpenHands)Vals · Code Migration 36.55%Vals · GPQA Diamond 94.44%Vals · MMLU Pro 90.22%Vals · ProofBench 48%
Enkrypt AI Safety Leaderboard ↗6 measuresread 2026-09-26
Enkrypt · Jailbreak risk 15.5%Enkrypt · Harmful content risk 3.9%Enkrypt · CBRN risk 36.2%Enkrypt · Toxicity risk 3.5%Enkrypt · Bias risk 67.7%Enkrypt · Insecure code risk 7.1%

Badge

PublicAI Index badge for Gemini 3.8 Flash[![PublicAI Index](https://publicai.io/model-index/badge?model=gemini-3-8-flash)](https://publicai.io/model-index/m/gemini-3-8-flash)