Request a pilot
‹ PublicAI Index

The LLM benchmark aggregator.

Claude Sonnet 5

Anthropic

Strongest in Safety (#1 of 337, on 1 of its 3 boards), weakest in Core abilities (#107 of 204). Above par in 30 of 34 scopes. Among the models it meets almost everywhere, it finishes behind Claude Fable 5 and Claude Fable 5.1 and ahead of Grok 4.5 and Gemini 3.1 Pro.

§ 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.

Safety61.5#1/337
Fairness70#5/300
Toxicity avoidance58.7#11/272
Jailbreak resistance62.8#12/272
Harm refusal59.6#13/300
Secure code62.1#45/274
Reasoning57.7#24/178
Mathematics57.7#21/140
Reasoning57.1#31/140
Science58.9#33/122
Expert reasoning59.4—/0✱
Agents60.7#25/268
Knowledge work63.9#22/178
Tool use59.4—/81✱
Web research58.3—/4✱
Workflow automation59.3—/0✱
Coding56.5#27/165
Code generation56.8#21/77
Agentic coding56.4#30/157
Professional55.9#29/168
Finance59.2#7/151
Legal55.2#40/151
Medical52.4#67/140
Knowledge56.6#29/138
Academic knowledge57.9#26/123
Factuality59—/0✱
Human preference63#46/342
Human preference63#46/342
Core abilities48.9#107/204
Data analysis47.1#40/57
Instruction following43#40/57
Language41.9#41/57
General intelligence56.3#54/204
Long context60.6—/0✱

§ 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 Claude Sonnet 5, left for the other.

Nemotron 3 Ultrawon 22 · lost 0
Gemini 2.0 Flashwon 22 · lost 0
GPT-5.4 Miniwon 22 · lost 0
Llama 4 Scout Instructwon 23 · lost 0
Command Awon 23 · lost 0
GPT-4o Miniwon 24 · lost 0
GPT-4.1 Nanowon 25 · lost 0
GPT-4.1 Miniwon 25 · lost 0
Grok 4.3won 27 · lost 1
DeepSeek V3won 24 · lost 1
Claude Haiku 4.5won 24 · lost 1
Gemma 4 31Bwon 24 · lost 1
O3 Miniwon 22 · lost 1
GPT-5.4 Nanowon 22 · lost 1
Gemini 3.5 Flash Litewon 21 · lost 1
GPT OSS 20Bwon 20 · lost 1
GPT-4owon 20 · lost 1
MiniMax M2.7won 20 · lost 1
Claude Sonnet 4won 23 · lost 2
Gemini 2.5 Flashwon 23 · lost 2
GPT OSS 120Bwon 22 · lost 2
GLM-4.6won 22 · lost 2
Grok 4.1 Fastwon 20 · lost 2
Qwen3.6 Pluswon 20 · lost 2
GLM-5.1won 20 · lost 2
Claude 3.7 Sonnetwon 19 · lost 2
DeepSeek V3.2won 22 · lost 3
GPT-6 Lunawon 22 · lost 3
Kimi K2won 21 · lost 3
Mimo V2.5 Prowon 20 · lost 3
MiniMax M3won 29 · lost 5
Inklingwon 29 · lost 5
GLM-5.2won 28 · lost 5
Claude Sonnet 4.5won 21 · lost 4
Grok 4won 20 · lost 4
Kimi K2.5won 20 · lost 4
DeepSeek V4 Flashwon 25 · lost 5
Claude Opus 4won 19 · lost 4
Kimi K2.6won 23 · lost 5
GPT-5.1won 18 · lost 4
Claude Sonnet 4.6won 21 · lost 5
Gemini 2.5 Prowon 20 · lost 5
GPT-5.6 Lunawon 27 · lost 7
GPT-4.1won 19 · lost 5
Qwen3 235B A22B Instructwon 19 · lost 5
GPT-5 Nanowon 17 · lost 5
O4 Miniwon 19 · lost 6
DeepSeek R1won 18 · lost 6
GPT-5won 18 · lost 6
GPT-5 Miniwon 18 · lost 7
O3won 17 · lost 8
GLM-5.3 Flashwon 19 · lost 9
Qwen3.7 Maxwon 19 · lost 9
Gemini 3.6 Flashwon 19 · lost 9
DeepSeek V4 Prowon 20 · lost 10
Gemini 3.5 Flashwon 19 · lost 10
Qwen3.8 27Bwon 17 · lost 9
GPT-5.2won 17 · lost 11
GPT-5.6 Terrawon 18 · lost 15
GPT-5.4won 15 · lost 13
Muse Spark 1.1won 15 · lost 13
Claude Opus 4.5won 12 · lost 11
Gemini 3.1 Prowon 12 · lost 11
Grok 4.5won 15 · lost 14
Gemini 3.8 Flashwon 14 · lost 14
Grok 4.7won 12 · lost 13
GLM-5.3won 11 · lost 13
Claude Opus 4.6won 11 · lost 13
GPT-6 Solwon 11 · lost 14
DeepSeek V4.1 Flashwon 10 · lost 13
Qwen3.8 Maxwon 12 · lost 16
Claude Opus 4.8won 11 · lost 17
Gemini 3 Prowon 8 · lost 13
Gemini 3.7 Flashwon 8 · lost 14
Grok 4.6won 10 · lost 18
Claude Opus 4.7won 9 · lost 19
GPT-5.6 Solwon 9 · lost 22
GPT-5.5won 8 · lost 20
Claude Opus 5won 8 · lost 22
Muse Spark 1.2won 5 · lost 16
Claude Opus 5.5won 5 · lost 20
Kimi K3won 4 · lost 20
Claude Fable 5.1won 0 · lost 22
Claude Fable 5won 0 · lost 23

§ 3 · Sources

Where the numbers come from

10 publications, 50 figures. Every one links to the page it was read from.

LMArena Text ↗1 measureread 2026-09-26
LMArena Text 1462
Artificial Analysis ↗1 measureread 2026-09-26
Artificial Analysis Intelligence Index 38
GDPval-AA ↗1 measureread 2026-09-26
GDPval-AA 1449
AA-Briefcase ↗1 measureread 2026-09-26
AA-Briefcase 1359
Terminal-Bench ↗1 measureread 2026-09-26
Terminal-Bench 12.4% (Claude Code)
LiveBench ↗8 measuresread 2026-09-26
LiveBench 76LiveBench · Reasoning 88.7LiveBench · Coding 80.7LiveBench · Agentic Coding 59.4LiveBench · Mathematics 92.9LiveBench · Data Analysis 71.7LiveBench · Language 75LiveBench · Instruction Following 63.9
SimpleBench ↗1 measureread 2026-09-26
SimpleBench 60.6%
Vals.ai ↗15 measuresread 2026-09-26
Vals · Legal Research Bench 41.83%Vals · LegalBench 83.92%Vals · Harvey Legal Agent Benchmark 5%Vals · Finance Agent 53.91%Vals · CorpFin 66.98%Vals · TaxEval 75.63%Vals · MortgageTax 70.03%Vals · MedCode 47.54%Vals · MedScribe 76.05%Vals · SWE-bench Verified 79.6% (Mini-SWE-agent)Vals · Vibe Code Bench 81.33% (OpenHands)Vals · Code Migration 44.39%Vals · GPQA Diamond 88.89%Vals · MMLU Pro 87.55%Vals · ProofBench 77%
Enkrypt AI Safety Leaderboard ↗6 measuresread 2026-09-26
Enkrypt · Jailbreak risk 2%Enkrypt · Harmful content risk 0%Enkrypt · CBRN risk 7.5%Enkrypt · Toxicity risk 0.2%Enkrypt · Bias risk 29.5%Enkrypt · Insecure code risk 8.9%
GDPVal-AA 1584tau3-Banking 37.3%Toolathlon Verified 71.6%Automation Bench Public 34.7%Apex-Agents (pass@1) 31.7%MCPMark 65.3%BrowseComp 84.7%Terminal-Bench 2.1 80.5%SciCode 53.6%Humanity's Last Exam (without tools) 41.3%GPQA Diamond 91.1%CritPt 16.9%AA-LCR 77%AA-Omniscience Accuracy 40%AA-Omniscience Non-Hallucination 61%

Badge

PublicAI Index badge for Claude Sonnet 5[![PublicAI Index](https://publicai.io/model-index/badge?model=claude-sonnet-5)](https://publicai.io/model-index/m/claude-sonnet-5)

✱ Placed by a one-off publication, not a board that re-ran the model. ✱✱ marks a figure the model’s own publisher printed, which counts for less again. Scores are 0–100 on the PublicAI Index scale, 50 = the average of the models each source lists. Scores belong to their publishers.

Snapshot 2026-09-26 · the full index · JSON API