“AI x Creativity" – The Weekly Sync

A whirlwind of breakthroughs, deals, and debates — here’s everything that mattered this week: Edition 8

Cahn Studios @ World Summit AI, San Francisco 2025. We were on-ground connecting with top researchers, developers, and creators shaping the future of AI. More insights coming soon!!!

This Week in AI (And What It Means for You)

Google Gemini 2.5 Models Now GA

Gemini 2.5 Pro and Gemini Flash are fully available for enterprise use. Plus, Gemini Flash-Lite enters preview — optimized for speed. Also, Launched open-source project building Full Stack AI Agents. You can run and customize it on Replit to build your own agent
New feature: Adjustable thinking sliders to balance speed vs. reasoning.

Open AI Drops GPT-4o Mini — Ultra Cheap & Fast

Open AI quietly launched GPT-4o Mini, a $0.03-per-million-token model. It's faster, cheaper, and smart enough for most day-to-day workflows.

The race to the bottom on price just hit warp speed.

Perplexity Teams Up With TIME

Perplexity becomes TIME’s official AI research partner. Original news meets AI summarization.

Great content still wins. Speed + source = future of news.

Anthropic Expands Claude 3 API Access

Claude 3 models now available to all developers via API. No waitlist.

AI tools are now collaborators, not just assistants.

Mistral Mixtral 8x22B Is Live

Mistral quietly launched Mixtral 8x22B, a mixture of experts model that balances top-tier performance with cost-efficiency.

Code snippets and multilingual Q&A are where it shines.

YouTube to Add AI-Generated Dubbing

YouTube Studio is testing built-in AI dubbing tools with ElevenLabs and Papercup.

Automatic translations could change global creator reach overnight.

Proactor — The Self-Initiating AI Agent

Proactor AI can take action before prompts. It’s context-aware, memory-powered, and hinting at the next generation of AI assistants.

Design velocity just doubled.

 MiniMax Launches 2 Flagship Models
 Hailuo 02: A powerful new video model, now ranked #2 globally

 M1: A 1M-token reasoning model built for long-form comprehension
At Cahn, we ran it locally. Here’s the real verdict:
Generates high-fidelity video content with strong temporal coherence and frame quality. And is now ranked #2 globally

Tools to Try This Week

  1. Ideogram – Next-gen image gen with wild typography control.

  2. Lamini – Enterprise-grade LLMs with extreme data privacy.

  3. Wordcab – Auto-summarize calls and meetings like a human.

  4. Tether – Build smart workflows using AI + API chains (Zapier killer).

Coming Up!!!!
Cahn Studios Masterclass – June 21 & 22- June28 &29
4-day AI crash course for creators + marketers.
🎟️ Save your spot — https://www.cahnstudios.com/webinar

👉 DM us on Instagram
📧 Or reply to this email — we’re listening.

Thought of the Week:
"AI isn’t just making us faster — it’s making us bolder."

Until next week,
Team Cahn

 Want the LTX API or Figma AI guide?
Email: [email protected]