AI PPT 뭐 결제할까? 클로드·캔바·젠스파크 싹 다 써본 솔직한 결론(한달에 PPT 수백장 뽑는 노하우 대방출)

Product(s) reviewed (AI tools for generating PPTs)

The video compares three AI services for creating lecture/PPT content—GenSpark, Claude, and Canva—and proposes a workflow using all three.

---

Key points & features mentioned (all unique)

1. AI-generated PPT end-to-end

   • Claims the PPTs shown are generated entirely by AI (lecture materials + exchanges/content).
   • Creator prints hundreds of PPT slides per month.

2. Core workflow: separate “research” vs “lecture writing”

   • Step 1 (materials research): GenSpark
     • Used to gather Korean-specific sources (Korean YouTube, Naver, blogs, news, communities).
     • Main claim: many AIs struggle with Korean data; GenSpark finds more Korean materials.
   • Step 2 (lecture exchange/write-up): Claude
     • Claude is said to produce more natural lecture exchanges (writing quality).
   • The creator uses a combination: GenSpark (materials) + Claude (writing).

3. “Don’t try to do everything in one tool”

   • Warning that attempting full end-to-end with only GenSpark leads to wasted credits and poorer results.

4. Font quality is critical

   • Main advice: use a Korean-optimized clean font: Pretendard.
   • The video provides a setup method from Noonnu (download and install Bold/Light/Regular only).
   • Goal: avoid “tacky” fonts that make the PPT look unprofessional.

5. PPT generation methods compared (how to turn text into slides)

   • Claude → PPT
     • Claim: adding the “MD”/lecture output to the Claude project can generate PPT.
     • Claude strength: replicate a design system accurately (brand style, colors, background cohesion).
     • Uses a free resource: getdin design.md (design system prompts).
   • GenSpark AI Slide
     • Highlights improved feature in version 5.0: a skills system.
     • You can select slide templates, save your “rules” as a reusable slide skill (style/color/layout preset).
     • Then future PPTs keep the same tone/style when reusing the skill.
     • Creator says previously GenSpark design wasn’t great, but skills improve consistency.
     • Downside implied: GenSpark is credit-based, and generating slides consumes credits.
   • Canva (spelled “Camba” in subtitles)
     • Strength: very pretty templates.
     • Workflow: choose “Presentation” templates, narrow search with keyword “stock lecture template”, then Edit with AI.
     • Paste the Claude lecture exchange and let Canva generate the deck.
     • Canva AI can fix layout issues like text overlapping; supports editing fonts/colors/images.

6. Pricing comparison (explicit numbers)

   • Canva Pro: 9,900 won/month
   • Claude Pro: $22/month (about 33,000 won incl. VAT)
   • GenSpark Plus: $24.99/month (about 37,000 won)
   • Statement: all roughly fall in 10,000–30,000 won range (though the converted numbers shown include ~33k–37k).

7. Overall recommendation

   • Creator personally uses all three monthly.
   • Best single pick (overall): GenSpark
     • Reason 1: can do both research + PPT creation in one go (two of the six demos).
     • Reason 2: quality is described as “overwhelming” and strong on both composition and design (thanks to agent-like behavior).
   • Tradeoff: GenSpark’s credit consumption can be a downside.

8. When to choose each tool (as decision guidance)

   • Choose Claude for lecture materials/e-books where brand tone matters.
   • Choose Canva if you need the most beautiful designs.
   • Choose GenSpark for standard presentation materials / “printing PPTs” workflows.

9. Quick actions recommended

   • Download Pretendard immediately from Noonnu.
   • Print one slide following the recommended workflow order:
     • GenSpark for materials → Claude for exchange

---

Pros (as stated/implied)

• GenSpark
  • Strong at finding Korean sources.
  • Can handle research + PPT generation efficiently.
  • With AI Slide skills, produces consistent styles across decks.
  • Creator reports fast completion (~30 minutes with proper tool selection).
• Claude
  • Produces natural lecture exchanges (writing quality).
  • Can replicate design systems well for consistent branding.
• Canva
  • Best template beauty among the three.
  • AI can fix layout issues automatically (e.g., overlapping text).
  • Easy edit loop via templates + “Edit with AI”.

---

Cons (as stated/implied)

• GenSpark
  • Credit-based; pulling/generating slides consumes credits.
  • If you try to do everything from start to finish within GenSpark, results/efficiency may suffer.
• Claude
  • No explicit downside stated in the subtitles (mainly positioned as writing-oriented).
• Canva
  • Not explicitly criticized; mainly positioned as template/design-focused rather than research/writing-first.

---

Comparisons made

• GenSpark vs Claude
  • GenSpark = research/material gathering (especially Korean content) + strong overall PPT output.
  • Claude = writing/natural lecture exchanges and design system replication.
• Canva vs GenSpark
  • Canva = prettier templates.
  • GenSpark = agent-like end-to-end workflow with consistent style features, but with credit costs.
• Tool pairing strategy
  • Best results come from splitting tasks: GenSpark (materials) → Claude (exchange) → Claude/GenSpark/Canva (slide build) depending on desired style and tone.

---

Overall verdict / recommendation

• Top recommendation: GenSpark (best overall for creating PPTs from Korean research through to slide decks).
• If you care most about writing natural lecture exchange and brand-consistent tone/design system prompting: use Claude.
• If you want prettiest templates and easy slide fixes: use Canva.

---

Speakers / contributors mentioned

• Primary speaker: “Saint”
  • Claims extensive hands-on experience and prints hundreds of PPTs monthly.
  • Gives the main comparisons, workflow, and pricing/recommendations.
• Minor mention
  • References “dev”/another person generally (e.g., “developer for 20 years”), but no separate product opinions are clearly attributed beyond Saint’s own guidance.