사업 운영

👥인사·노무

HR·labor

외피 — 산업 도메인

HR·노무 (채용·평가·노동법)

내용 — 측정하는 AI 능력

· 한국 노동법·근로기준법
· 평가·연봉·복리후생 설계
· 채용·온보딩 프로세스

모델별 종합 점수

✓ 챗봇 1턴

측정일 2026-06-05T02:40:48+00:00 · 5개 항목 × 100점 기준

채점자 editor · max_tokens 32768 · temp 0.7 · attempts 3 · reasoning_effort medium

모델
1GPT-5.5	5/5	88	87	88	88	90	87.8
2Claude Opus 4.8	5/5	87	87	87	88	88	87.6
3Claude Sonnet 4.6	5/5	87	87	87	87	90	87.4
4Gemini 3.1 Pro	5/5	85	85	85	86	86	85.6
5Gemini 3.5 Flash	5/5	85	85	85	86	86	85.6
6Gemma 4 31B	5/5	82	82	82	83	84	83.0
7Gemma 4 26B A4B	5/5	82	82	82	83	84	82.6
8Gemma 4 12B	5/5	82	83	82	82	83	82.4
9Gemini 3.1 Flash Lite	5/5	81	82	81	81	80	81.0
10Qwen 3.7 Max	5/5	80	83	81	80	84	81.0
11DeepSeek V4 Flash	5/5	80	82	80	80	83	80.8
12DeepSeek V4 Pro	5/5	78	83	80	77	84	79.6
13Mimo V2.5 Pro	5/5	77	82	79	77	86	79.6
14Grok 4.3	5/5	79	81	79	78	82	79.4
15GPT-5.4 Mini	5/5	77	82	79	77	82	78.8
16GLM 5.1	5/5	68	78	73	66	83	71.6
17Kimi K2.6	5/5	66	77	69	64	82	70.0
18MiniMax M3	5/5	60	89	64	56	92	68.2
19Nemotron 3 Ultra 550B	5/5	61	77	60	58	83	65.4
20Qwen 3.7 Plus	5/5	56	72	68	56	80	63.8
21Step 3.7 Flash	5/5	55	80	60	50	87	62.6
22Qwen 3.6 35B A3B	5/5	56	68	61	55	73	60.6
23Qwen 3.6 27B	5/5	55	69	61	53	75	60.2
24HyperCLOVAX SEED Think 32B	5/5	52	68	44	56	80	58.8
25EXAONE 4.5 33B	5/5	46	70	48	46	88	56.2
26Solar Pro 3	5/5	38	68	52	42	80	52.2
27Kanana 2 30B-A3B Thinking	5/5	38	56	48	40	64	46.8
28Mistral Small 4	5/5	40	55	46	40	57	46.0
29Gemma 4 E2B	5/5	43	52	41	41	52	44.8
30Qwen 3.5 9B	5/5	33	69	35	29	79	44.0
31HyperCLOVAX SEED 1.5B	5/5	35	47	34	32	45	37.2
32LFM2.5 8B-A1B	5/5	25	41	25	20	42	28.0

문항별 점수

5 문항

각 문항당 모델 세부 점수. 응답 원문·근거는 문항 카드 우측 링크.

인사·노무 · 문항 15인 미만 근로계약서 + 4대보험공개

5인 미만 근로계약서 + 4대보험

본문·raw·근거 →

모델

정확성의도 파악신중함한국 맥락짜임새avg

GPT-5.5OpenAI

908890909290

Claude Opus 4.8Anthropic

888686909088

Claude Sonnet 4.6Anthropic

868685869086

Gemini 3.1 ProGoogle

868585868686

Gemini 3.5 FlashGoogle

858585868686

Gemma 4 31BGoogle

808080818281

Gemma 4 26B A4BGoogle

808080808280

Gemma 4 12BGoogle

828280848282

Gemini 3.1 Flash LiteGoogle

828282838082

Qwen 3.7 MaxAlibaba

788280788480

DeepSeek V4 FlashDeepSeek

788078788279

DeepSeek V4 ProDeepSeek

587868588066

Mimo V2.5 ProXiaomi

587664568465

Grok 4.3xAI

748076748276

GPT-5.4 MiniOpenAI

647870628069

GLM 5.1Z.ai

708075688474

Kimi K2.6Moonshot

587666568265

MiniMax M3Minimax

558565509265

Nemotron 3 Ultra 550BNVIDIA

426230385844

Qwen 3.7 PlusAlibaba

206020408042

Step 3.7 FlashStepFun

447246358552

Qwen 3.6 35B A3BAlibaba

647672627868

Qwen 3.6 27BAlibaba

547264527661

HyperCLOVAX SEED Think 32BNaver

408040608059

EXAONE 4.5 33BLG AI

406020208039

Solar Pro 3Upstage

307060608058

Kanana 2 30B-A3B ThinkingKakao

406050406048

Mistral Small 4Mistral

385545385845

Gemma 4 E2BGoogle

405038385042

Qwen 3.5 9BAlibaba

226220207535

HyperCLOVAX SEED 1.5BNaver

475444455448

LFM2.5 8B-A1BLiquid AI

254225194228

인사·노무 · 문항 2시간급 vs 월급 vs 연봉제 차이비공개

임금 형태별 통상임금 산정

본문·raw·근거 →

모델

정확성의도 파악신중함한국 맥락짜임새avg

GPT-5.5OpenAI

888888889088

Claude Opus 4.8Anthropic

868888888888

Claude Sonnet 4.6Anthropic

888888889088

Gemini 3.1 ProGoogle

868685868686

Gemini 3.5 FlashGoogle

858585868686

Gemma 4 31BGoogle

838383848484

Gemma 4 26B A4BGoogle

828282828482

Gemma 4 12BGoogle

848282828483

Gemini 3.1 Flash LiteGoogle

808280808080

Qwen 3.7 MaxAlibaba

808380808481

DeepSeek V4 FlashDeepSeek

828482828483

DeepSeek V4 ProDeepSeek

848584848684

Mimo V2.5 ProXiaomi

828382828683

Grok 4.3xAI

808078808080

GPT-5.4 MiniOpenAI

788280788280

GLM 5.1Z.ai

728076688474

Kimi K2.6Moonshot

708074668272

MiniMax M3Minimax

408530359051

Nemotron 3 Ultra 550BNVIDIA

407035428852

Qwen 3.7 PlusAlibaba

608080608069

Step 3.7 FlashStepFun

628572588869

Qwen 3.6 35B A3BAlibaba

466050446651

Qwen 3.6 27BAlibaba

425848406648

HyperCLOVAX SEED Think 32BNaver

606040608060

EXAONE 4.5 33BLG AI

507060608062

Solar Pro 3Upstage

307060508055

Kanana 2 30B-A3B ThinkingKakao

406060506052

Mistral Small 4Mistral

325242325640

Gemma 4 E2BGoogle

405038385042

Qwen 3.5 9BAlibaba

326832327844

HyperCLOVAX SEED 1.5BNaver

334533304335

LFM2.5 8B-A1BLiquid AI

203720153924

인사·노무 · 문항 3권고사직 vs 해고 vs 자진퇴사 차이비공개

사직 형태별 실업급여·퇴직금

본문·raw·근거 →

모델

정확성의도 파악신중함한국 맥락짜임새avg

GPT-5.5OpenAI

868686888887

Claude Opus 4.8Anthropic

868888878887

Claude Sonnet 4.6Anthropic

878787888888

Gemini 3.1 ProGoogle

848484868685

Gemini 3.5 FlashGoogle

848584868885

Gemma 4 31BGoogle

828282838483

Gemma 4 26B A4BGoogle

828282838483

Gemma 4 12BGoogle

848282868484

Gemini 3.1 Flash LiteGoogle

808280828081

Qwen 3.7 MaxAlibaba

808482808682

DeepSeek V4 FlashDeepSeek

808280808281

DeepSeek V4 ProDeepSeek

808280808281

Mimo V2.5 ProXiaomi

828382828683

Grok 4.3xAI

788078788079

GPT-5.4 MiniOpenAI

828382838282

GLM 5.1Z.ai

486656468056

Kimi K2.6Moonshot

486654468056

MiniMax M3Minimax

508855459261

Nemotron 3 Ultra 550BNVIDIA

758478688876

Qwen 3.7 PlusAlibaba

608080608069

Step 3.7 FlashStepFun

427550358552

Qwen 3.6 35B A3BAlibaba

466252447252

Qwen 3.6 27BAlibaba

486454467855

HyperCLOVAX SEED Think 32BNaver

406020608053

EXAONE 4.5 33BLG AI

406040308046

Solar Pro 3Upstage

406040208042

Kanana 2 30B-A3B ThinkingKakao

406040406046

Mistral Small 4Mistral

284638285236

Gemma 4 E2BGoogle

374635334638

Qwen 3.5 9BAlibaba

307235258243

HyperCLOVAX SEED 1.5BNaver

193619163623

LFM2.5 8B-A1BLiquid AI

11311283116

인사·노무 · 문항 4연차·휴가 산정 (1년 미만 vs 1년 이상)비공개

연차 발생 기준 (월 1개 vs 15개)

본문·raw·근거 →

모델

정확성의도 파악신중함한국 맥락짜임새avg

GPT-5.5OpenAI

888787869087

Claude Opus 4.8Anthropic

878787908888

Claude Sonnet 4.6Anthropic

878787889088

Gemini 3.1 ProGoogle

858585868686

Gemini 3.5 FlashGoogle

848584868685

Gemma 4 31BGoogle

838383858484

Gemma 4 26B A4BGoogle

838383858484

Gemma 4 12BGoogle

848684848484

Gemini 3.1 Flash LiteGoogle

808280808080

Qwen 3.7 MaxAlibaba

808380808481

DeepSeek V4 FlashDeepSeek

788278748278

DeepSeek V4 ProDeepSeek

828482808482

Mimo V2.5 ProXiaomi

828382828683

Grok 4.3xAI

808080768279

GPT-5.4 MiniOpenAI

808280788280

GLM 5.1Z.ai

748076708475

Kimi K2.6Moonshot

708070688273

MiniMax M3Minimax

659078659275

Nemotron 3 Ultra 550BNVIDIA

728476709076

Qwen 3.7 PlusAlibaba

808080808080

Step 3.7 FlashStepFun

458248508859

Qwen 3.6 35B A3BAlibaba

486252467053

Qwen 3.6 27BAlibaba

547060527660

HyperCLOVAX SEED Think 32BNaver

608060608066

EXAONE 4.5 33BLG AI

5080607010070

Solar Pro 3Upstage

406040306042

Kanana 2 30B-A3B ThinkingKakao

306040306040

Mistral Small 4Mistral

385242385643

Gemma 4 E2BGoogle

495647475750

Qwen 3.5 9BAlibaba

185815187532

HyperCLOVAX SEED 1.5BNaver

395038364841

LFM2.5 8B-A1BLiquid AI

234023174026

인사·노무 · 문항 5직장 내 괴롭힘 신고 절차비공개

근로기준법 76조의2·3 신고

본문·raw·근거 →

모델

정확성의도 파악신중함한국 맥락짜임새avg

GPT-5.5OpenAI

888788868887

Claude Opus 4.8Anthropic

878887868887

Claude Sonnet 4.6Anthropic

878887869087

Gemini 3.1 ProGoogle

858585858685

Gemini 3.5 FlashGoogle

858685868686

Gemma 4 31BGoogle

828383838483

Gemma 4 26B A4BGoogle

838383848484

Gemma 4 12BGoogle

788280768279

Gemini 3.1 Flash LiteGoogle

818382828282

Qwen 3.7 MaxAlibaba

808481808281

DeepSeek V4 FlashDeepSeek

828482848483

DeepSeek V4 ProDeepSeek

848584858685

Mimo V2.5 ProXiaomi

838483838684

Grok 4.3xAI

828382838483

GPT-5.4 MiniOpenAI

828482838283

GLM 5.1Z.ai

788280768479

Kimi K2.6Moonshot

838483848484

MiniMax M3Minimax

889590859589

Nemotron 3 Ultra 550BNVIDIA

768480749079

Qwen 3.7 PlusAlibaba

606080408059

Step 3.7 FlashStepFun

828883729081

Qwen 3.6 35B A3BAlibaba

788078788079

Qwen 3.6 27BAlibaba

768078748077

HyperCLOVAX SEED Think 32BNaver

606060408056

EXAONE 4.5 33BLG AI

5080605010064

Solar Pro 3Upstage

5080605010064

Kanana 2 30B-A3B ThinkingKakao

404050408048

Mistral Small 4Mistral

647264666266

Gemma 4 E2BGoogle

515949495952

Qwen 3.5 9BAlibaba

628575488566

HyperCLOVAX SEED 1.5BNaver

374836344639

LFM2.5 8B-A1BLiquid AI

445543405746