YouTube Video
Get the Agent OS & Jev AI Masterclass 👉 https://www.skool.com/ai-profit-lab-7462/about
Want to make money and save time with AI? Join here: https://www.skool.com/ai-profit-lab-7462/about
Get a FREE AI Automation Strategy Session: https://juliangoldieaiautomation.com/
Get a FREE AI SEO Strategy Session: https://go.juliangoldie.com/strategy-session?utm=julian
Stop Wasting Claude & GPT-6 Astra Tokens: Cut Token Use Fast with Jev AI
This episode explains how to reduce Claude and GPT-6 Astra token usage by offloading “tiny choice” tasks to Jev AI, a small decision model from Typesafe AI released September 15 that selects from allowed answers and returns a confidence score without generating text. It defines tokens, notes that large models reread entire chat histories every message, and shows examples where short prompts triggered huge token reads. The script compares Jev against Claude Haiku, Claude Sonnet 5, and GPT-6 Astra on an outfit-selection workflow, highlighting faster response times and lower cost, and demonstrates similar use cases like voice browser control and email triage. It outlines three implementation steps: identify quiz-like decisions, integrate Jev via API (e.g., through OpenRouter) into agents, and use confidence thresholds to decide when to escalate to a larger model.
00:00 Stop Wasting Tokens
00:47 Tokens Explained Fast
01:08 The Hidden Chat Tax
01:58 Why Big Models Waste
02:25 Meet Jev AI
03:43 Outfit Picker Race
04:57 Real Time Choice Systems
06:20 Three Step Setup
07:57 Confidence Thresholds
08:20 Costs Speed and Wrap Up
09:14 Final Call to Action
Want to make money and save time with AI? Join here: https://www.skool.com/ai-profit-lab-7462/about
Get a FREE AI Automation Strategy Session: https://juliangoldieaiautomation.com/
Get a FREE AI SEO Strategy Session: https://go.juliangoldie.com/strategy-session?utm=julian
Stop Wasting Claude & GPT-6 Astra Tokens: Cut Token Use Fast with Jev AI
This episode explains how to reduce Claude and GPT-6 Astra token usage by offloading “tiny choice” tasks to Jev AI, a small decision model from Typesafe AI released September 15 that selects from allowed answers and returns a confidence score without generating text. It defines tokens, notes that large models reread entire chat histories every message, and shows examples where short prompts triggered huge token reads. The script compares Jev against Claude Haiku, Claude Sonnet 5, and GPT-6 Astra on an outfit-selection workflow, highlighting faster response times and lower cost, and demonstrates similar use cases like voice browser control and email triage. It outlines three implementation steps: identify quiz-like decisions, integrate Jev via API (e.g., through OpenRouter) into agents, and use confidence thresholds to decide when to escalate to a larger model.
00:00 Stop Wasting Tokens
00:47 Tokens Explained Fast
01:08 The Hidden Chat Tax
01:58 Why Big Models Waste
02:25 Meet Jev AI
03:43 Outfit Picker Race
04:57 Real Time Choice Systems
06:20 Three Step Setup
07:57 Confidence Thresholds
08:20 Costs Speed and Wrap Up
09:14 Final Call to Action
SOCIAL SHARE CARD GENERATOR