GPT 5.6 Luna Just Got 80% Cheaper: Do This Now

⚙️ Book an AI project call → https://www.genial-agency.com/en#book 📬 Read my weekly newsletter → https://substack.com/@migueltorrez 🤝 Connect on LinkedIn → https://www.linkedin.com/in/miguel-torrez/

See how I configure GPT 5.6 Sol to delegate tasks to Luna subagents in Codex or ChatGPT Work. I explain the price drop discussed in the video, where Luna falls short, and the prompt for maximum reasoning with fast mode.

⏱️ TIMESTAMPS:

0:00 GPT 5.6 Luna just got 80% cheaper 0:35 Benchmarks: Luna vs DeepSeek V4 Pro 1:14 The nuances: Sol vs Luna 2:15 Where Luna still falls short (design & UX) 2:28 The power move: Sol main driver + Luna subagents 3:15 The setup prompt (Codex & ChatGPT Work) 3:54 And we're back: verifying the config 4:08 The Max reasoning mode gotcha 4:24 Wrap-up: 20x cheaper subagents

🧩 PROMPT FROM THE VIDEO:

Hey ChatGPT, whenever you work with subagents, I want you to run them as Luna agents with the highest reasoning setting possible and fast mode. Can you set that up for me in this project?

Note: Luna supports a "Max" reasoning mode that the settings UI doesn't show (it stops at Extra High), so ask ChatGPT or Codex for max mode explicitly.

🛠️ TOOLS MENTIONED:

• ChatGPT: https://openai.com • Codex: https://chatgpt.com/codex

🔗 MORE FROM ME: 𝕏 X: https://x.com/migueltorrez 📸 Instagram: https://www.instagram.com/migueltorrez.ai/ 🎵 TikTok: https://www.tiktok.com/@migueltorrez.ai 🇫🇷 Luwai (French): https://luwai.fr 👤 Personal: https://migueltorrez.ai

💬 DROP A COMMENT:

Are you already using Luna subagents, or is Sol still doing all your heavy lifting? Let me know what setup you're running.

#GPT56Luna #ChatGPT #Codex