사업 운영

📣마케팅·콘텐츠

Marketing·content

외피 — 산업 도메인

마케팅·콘텐츠·브랜드

내용 — 측정하는 AI 능력

· 카피·후크 생성
· 채널별 톤 (X·LinkedIn·네이버·유튜브)
· 한국 시장 인사이트·세그먼트

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
1MiniMax M3	5/5	87	92	91	90	93	91.2
2GPT-5.5	5/5	87	86	87	87	88	87.4
3Claude Opus 4.8	5/5	87	87	87	86	88	87.2
4Claude Sonnet 4.6	5/5	86	86	86	86	87	86.2
5Gemini 3.1 Pro	5/5	85	85	85	85	86	85.2
6Nemotron 3 Ultra 550B	5/5	74	88	82	83	90	85.0
7Kimi K2.6	5/5	84	84	84	84	85	84.6
8Gemini 3.5 Flash	5/5	83	83	83	84	84	83.8
9GLM 5.1	5/5	83	83	83	83	85	83.8
10GPT-5.4 Mini	5/5	83	83	83	83	84	83.4
11Mimo V2.5 Pro	5/5	84	84	84	84	79	82.6
12Qwen 3.7 Max	5/5	82	84	82	81	83	82.6
13Step 3.7 Flash	5/5	70	89	80	78	87	82.6
14Gemma 4 12B	5/5	82	85	82	80	84	82.4
15Gemma 4 31B	5/5	80	81	81	82	82	81.6
16Qwen 3.7 Plus	5/5	82	83	81	81	82	81.6
17DeepSeek V4 Pro	5/5	81	82	82	81	81	81.4
18DeepSeek V4 Flash	5/5	80	81	80	81	82	81.2
19Gemma 4 26B A4B	5/5	80	80	81	82	80	80.8
20Qwen 3.6 35B A3B	5/5	79	81	80	79	80	79.8
21Gemini 3.1 Flash Lite	5/5	80	80	80	80	78	79.4
22Grok 4.3	5/5	79	80	80	80	78	79.0
23Qwen 3.5 9B	5/5	72	83	78	70	81	77.0
24Qwen 3.6 27B	5/5	77	79	78	74	78	76.8
25EXAONE 4.5 33B	5/5	60	80	66	66	86	74.2
26Solar Pro 3	5/5	58	72	66	66	86	72.4
27HyperCLOVAX SEED Think 32B	5/5	60	72	56	72	68	68.0
28Gemma 4 E2B	5/5	57	65	54	62	69	63.2
29Mistral Small 4	5/5	58	67	61	55	67	61.8
30Kanana 2 30B-A3B Thinking	5/5	44	64	58	46	64	56.0
31LFM2.5 8B-A1B	5/5	41	49	39	41	51	45.4
32HyperCLOVAX SEED 1.5B	5/5	24	34	24	26	34	29.6

문항별 점수

5 문항

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

마케팅·콘텐츠 · 문항 11인 SaaS 30일 콘텐츠 캘린더 (네이버 SEO + 카카오)공개

1인 SaaS 30일 콘텐츠 캘린더 (네이버 SEO + 카카오)

본문·raw·근거 →

모델

정확성의도 파악신중함한국 맥락짜임새avg

MiniMax M3Minimax

889088909591

GPT-5.5OpenAI

888686889289

Claude Opus 4.8Anthropic

868686868887

Claude Sonnet 4.6Anthropic

868584868886

Gemini 3.1 ProGoogle

858484858685

Nemotron 3 Ultra 550BNVIDIA

658874759081

Kimi K2.6Moonshot

838482828684

Gemini 3.5 FlashGoogle

838383848484

GLM 5.1Z.ai

848484858685

GPT-5.4 MiniOpenAI

848484848685

Mimo V2.5 ProXiaomi

848484848685

Qwen 3.7 MaxAlibaba

808280768280

Step 3.7 FlashStepFun

628878759081

Gemma 4 12BGoogle

808482788682

Gemma 4 31BGoogle

828282838483

Qwen 3.7 PlusAlibaba

808080788280

DeepSeek V4 ProDeepSeek

788080787477

DeepSeek V4 FlashDeepSeek

828282828483

Gemma 4 26B A4BGoogle

808080827679

Qwen 3.6 35B A3BAlibaba

767878767275

Gemini 3.1 Flash LiteGoogle

788078787076

Grok 4.3xAI

767878766874

Qwen 3.5 9BAlibaba

708072728577

Qwen 3.6 27BAlibaba

768078708076

EXAONE 4.5 33BLG AI

608060609073

Solar Pro 3Upstage

6080607010079

HyperCLOVAX SEED Think 32BNaver

608060806070

Gemma 4 E2BGoogle

576452577162

Mistral Small 4Mistral

606862507062

Kanana 2 30B-A3B ThinkingKakao

406050305045

LFM2.5 8B-A1BLiquid AI

455243475650

HyperCLOVAX SEED 1.5BNaver

142817172822

마케팅·콘텐츠 · 문항 2LinkedIn 한국 톤 — 미국식이 왜 안 먹히나 분석비공개

한국 LinkedIn 톤 분석

본문·raw·근거 →

모델

정확성의도 파악신중함한국 맥락짜임새avg

MiniMax M3Minimax

859588909291

GPT-5.5OpenAI

868886868686

Claude Opus 4.8Anthropic

868886868887

Claude Sonnet 4.6Anthropic

858885868686

Gemini 3.1 ProGoogle

848684858585

Nemotron 3 Ultra 550BNVIDIA

809086869087

Kimi K2.6Moonshot

828582838484

Gemini 3.5 FlashGoogle

828482838383

GLM 5.1Z.ai

828582828483

GPT-5.4 MiniOpenAI

828482828282

Mimo V2.5 ProXiaomi

828582838484

Qwen 3.7 MaxAlibaba

828582828483

Step 3.7 FlashStepFun

759082828885

Gemma 4 12BGoogle

808682788482

Gemma 4 31BGoogle

808280828282

Qwen 3.7 PlusAlibaba

828580828483

DeepSeek V4 ProDeepSeek

808280808281

DeepSeek V4 FlashDeepSeek

808280808281

Gemma 4 26B A4BGoogle

808280828282

Qwen 3.6 35B A3BAlibaba

808280808281

Gemini 3.1 Flash LiteGoogle

788078787878

Grok 4.3xAI

808280808080

Qwen 3.5 9BAlibaba

688070455861

Qwen 3.6 27BAlibaba

788078767878

EXAONE 4.5 33BLG AI

608060709076

Solar Pro 3Upstage

608060709076

HyperCLOVAX SEED Think 32BNaver

608060808076

Gemma 4 E2BGoogle

637459707470

Mistral Small 4Mistral

626462526460

Kanana 2 30B-A3B ThinkingKakao

507060607064

LFM2.5 8B-A1BLiquid AI

515748526155

HyperCLOVAX SEED 1.5BNaver

263726283732

마케팅·콘텐츠 · 문항 33채널 카피 — 정밀 제약(자수·문장수·금칙어) 동시충족 산출비공개

3채널 정밀제약 카피

본문·raw·근거 →

모델

정확성의도 파악신중함한국 맥락짜임새avg

MiniMax M3Minimax

889290889691

GPT-5.5OpenAI

888687869088

Claude Opus 4.8Anthropic

878688868887

Claude Sonnet 4.6Anthropic

878687869087

Gemini 3.1 ProGoogle

858585858786

Nemotron 3 Ultra 550BNVIDIA

858888869088

Kimi K2.6Moonshot

848484848685

Gemini 3.5 FlashGoogle

848484858685

GLM 5.1Z.ai

848484848685

GPT-5.4 MiniOpenAI

848484848685

Mimo V2.5 ProXiaomi

858585858886

Qwen 3.7 MaxAlibaba

828480828483

Step 3.7 FlashStepFun

808885758281

Gemma 4 12BGoogle

828478808482

Gemma 4 31BGoogle

808080828281

Qwen 3.7 PlusAlibaba

808276808080

DeepSeek V4 ProDeepSeek

828282828483

DeepSeek V4 FlashDeepSeek

788074807879

Gemma 4 26B A4BGoogle

808080828281

Qwen 3.6 35B A3BAlibaba

828282828483

Gemini 3.1 Flash LiteGoogle

828282828483

Grok 4.3xAI

808080808080

Qwen 3.5 9BAlibaba

708580759082

Qwen 3.6 27BAlibaba

747672707473

EXAONE 4.5 33BLG AI

608060609073

Solar Pro 3Upstage

504060506052

HyperCLOVAX SEED Think 32BNaver

606040604052

Gemma 4 E2BGoogle

525750576458

Mistral Small 4Mistral

667062666466

Kanana 2 30B-A3B ThinkingKakao

405050306046

LFM2.5 8B-A1BLiquid AI

243324223328

HyperCLOVAX SEED 1.5BNaver

8151081512

마케팅·콘텐츠 · 문항 4콜드아웃리치 — 미국발 거짓 수치 전제 환각저항비공개

한국 콜드아웃 환각저항

본문·raw·근거 →

모델

정확성의도 파악신중함한국 맥락짜임새avg

MiniMax M3Minimax

909295928891

GPT-5.5OpenAI

888687908688

Claude Opus 4.8Anthropic

888888888888

Claude Sonnet 4.6Anthropic

878586888586

Gemini 3.1 ProGoogle

858485868585

Nemotron 3 Ultra 550BNVIDIA

608870809082

Kimi K2.6Moonshot

868586888586

Gemini 3.5 FlashGoogle

838383858384

GLM 5.1Z.ai

828282848483

GPT-5.4 MiniOpenAI

828282848283

Mimo V2.5 ProXiaomi

858485878585

Qwen 3.7 MaxAlibaba

858485868485

Step 3.7 FlashStepFun

629072809082

Gemma 4 12BGoogle

868686848284

Gemma 4 31BGoogle

808080828081

Qwen 3.7 PlusAlibaba

848484828283

DeepSeek V4 ProDeepSeek

828282838383

DeepSeek V4 FlashDeepSeek

808080828281

Gemma 4 26B A4BGoogle

808080828081

Qwen 3.6 35B A3BAlibaba

768078788079

Gemini 3.1 Flash LiteGoogle

808080827880

Grok 4.3xAI

808080828081

Qwen 3.5 9BAlibaba

828585828584

Qwen 3.6 27BAlibaba

788078787878

EXAONE 4.5 33BLG AI

608080708075

Solar Pro 3Upstage

608080709078

HyperCLOVAX SEED Think 32BNaver

808080808080

Gemma 4 E2BGoogle

596656636964

Mistral Small 4Mistral

305538345845

Kanana 2 30B-A3B ThinkingKakao

507070607065

LFM2.5 8B-A1BLiquid AI

475546495752

HyperCLOVAX SEED 1.5BNaver

445142465148

마케팅·콘텐츠 · 문항 5메타 광고 저예산 — 과장 기대 보정비공개

저예산 광고 기대 보정

본문·raw·근거 →

모델

정확성의도 파악신중함한국 맥락짜임새avg

MiniMax M3Minimax

859095909592

GPT-5.5OpenAI

868588858686

Claude Opus 4.8Anthropic

868688868887

Claude Sonnet 4.6Anthropic

858588858686

Gemini 3.1 ProGoogle

848486858585

Nemotron 3 Ultra 550BNVIDIA

808890869087

Kimi K2.6Moonshot

838386838484

Gemini 3.5 FlashGoogle

828385838383

GLM 5.1Z.ai

828285828483

GPT-5.4 MiniOpenAI

828285828282

Mimo V2.5 ProXiaomi

828285825273

Qwen 3.7 MaxAlibaba

828385808282

Step 3.7 FlashStepFun

709085808784

Gemma 4 12BGoogle

808484808482

Gemma 4 31BGoogle

808083828081

Qwen 3.7 PlusAlibaba

828285828282

DeepSeek V4 ProDeepSeek

828285828383

DeepSeek V4 FlashDeepSeek

808184818282

Gemma 4 26B A4BGoogle

808083828081

Qwen 3.6 35B A3BAlibaba

808184808281

Gemini 3.1 Flash LiteGoogle

808082807880

Grok 4.3xAI

808083808080

Qwen 3.5 9BAlibaba

728585758581

Qwen 3.6 27BAlibaba

788082787879

EXAONE 4.5 33BLG AI

608070708074

Solar Pro 3Upstage

608070709077

HyperCLOVAX SEED Think 32BNaver

406040608062

Gemma 4 E2BGoogle

566353616762

Mistral Small 4Mistral

747680727876

Kanana 2 30B-A3B ThinkingKakao

407060507060

LFM2.5 8B-A1BLiquid AI

374635374842

HyperCLOVAX SEED 1.5BNaver

293827303934