Hey Radarians, this week is about AI getting cheaper and more competitive.
Lower model prices, hardware battles, smarter follow-ups, health research, and a fresh AI side hustle.
Let’s dive in.
What’s Inside 📡
Radar News Brief
Steal This Workflow
AI Research
Tools Worth Watching
Prompt Radar
AI Side Hustle Idea
Learn AI with This Course
Radar News Brief
⚡ DeepSeek V4 Flash drives frontier AI prices sharply lower
DeepSeek priced V4 Flash at $0.14 per million input tokens and $0.28 for output, making advanced AI dramatically cheaper despite massive infrastructure spending by rivals
Independent analysis ranks it as the cheapest recognized model, while latest releases across the U.S. and China keep intensifying competition over price, speed and capability for global AI buyers
As model gaps narrow, businesses can route each task by cost, speed and skill, pressuring premium labs as frontier intelligence becomes easier to switch and scale
🍎 OpenAI publicly challenges Apple’s trade secrets lawsuit
OpenAI accused Apple of filing a careless and personal lawsuit, publishing emails and messages to dispute claims that former Apple employees brought confidential information into its hardware
Apple says Chang Liu and Tang Tan misused trade secrets to support OpenAI’s consumer device plans and is seeking an injunction limiting possession, use or disclosure while litigation continues
OpenAI says it neither has nor wants Apple’s secrets, arguing Apple failed to control sensitive employee info and misrepresented communications as competition over future AI hardware
More News on the Radar
Anthropic’s CEO says some top AI hires may be joining for huge pay, not just mission, as salary alone can’t retain talent
Qwen 3.8-Max is a 2.4T-parameter model improving coding, research, work, and long-horizon tasks; open weights next week
White House invites OpenAI, Anthropic, Meta, and Google to review latest AI cybersecurity testing rules after recent breaches
⚡ Steal This Workflow
Build a Smart Follow-Up Agent
A lead is most valuable when their interest is still fresh. Build an agent that responds personally and keeps the conversation moving.
Capture the inquiry
Connect your website form, Typeform, or Webflow form to n8n. Collect the lead’s name, work email, company website, request, and any answers that reveal what they need.
Research the business
Send the company name and website to Tavily or Perplexity.
Ask it to find useful context such as what the company sells, who it serves, recent developments, and problems your service could realistically solve.
Write the first email
Give the form answers and research to ChatGPT or Claude.
Agent instruction:
Write a short follow-up email for this inbound lead. Mention their specific request and one relevant detail about their business. Explain the most logical next step and include a clear call to action. Use supplied information and avoid exaggerated claims.Save the email as a Gmail draft for approval or send it automatically when the lead meets your qualification rules.
Continue when necessary
Store the Gmail thread ID, wait two or three business days, and check whether the lead replied.
Reply received → update the CRM and alert sales.
No reply → draft a brief follow-up in the same email thread.
🔎 AI RESEARCH: An AI Symptom Checker Was Preferred Over Doctors’ Lists
What researchers studied:
Google tested SymptomAI with 13,917 people to see how well it could ask about symptoms and suggest possible health conditions.
What they found:
Medical experts preferred SymptomAI’s lists over lists created by other doctors in more than half of the cases.
The AI performed better when it asked patients follow-up questions. However, it is still an experimental research tool and cannot replace a real medical examination.
Why it matters:
AI could help people understand their symptoms and prepare for a doctor’s visit, especially when healthcare is difficult to obtain.
Source: Google Research, July 2026
🛠️ Tools Worth Watching
ChatCut: Timeline-based AI video editor with advanced editing tools
Paradigm: Personalized AI learning paths adapting to your progress
Canva Code 2: Design websites and apps with Canva's AI tools
📝 Prompt Radar: Write Persuasive Business Proposals
What it does: Builds a proposal around the client’s real needs.
Act as a business proposal strategist. Using the information provided, create a persuasive proposal focused on the client’s problem, desired outcome, and decision criteria. Begin with an executive summary that shows a clear understanding of the situation. Then explain the current problem, proposed solution, scope, deliverables, process, timeline, responsibilities, success measures, assumptions, exclusions, risks, and next steps. Connect every feature of the solution to a practical client benefit. Avoid exaggerated promises and unsupported claims. Make pricing easy to understand if pricing is provided. End with a clear approval process and CTA. Client details: [PASTE]. Proposed service: [PASTE]. Evidence or proof: [PASTE].Best for:
Freelancers
Agencies
Consultants
💼 AI Side Hustle Idea: AI Search Visibility Audit
Most companies don’t need “more SEO content.” They need to know whether AI tools mention, recommend, or cite them when potential customers ask relevant questions.
The problem:
A company may rank on Google but remain invisible—or appear inaccurately—across ChatGPT, Perplexity, Gemini, and AI search results.
Your AI service:
Test real buyer questions across major AI platforms, track brand mentions and citations, analyze which competitors appear, and recommend improvements to the company’s priority pages.
Simple workflow:
Buyer questions → test AI platforms → record mentions → inspect cited sources → improve key pages → retest visibility
Best clients:
SaaS companies, agencies, consultants, e-commerce brands, local businesses, professional firms, and B2B service providers.
What you can charge for:
One-time audit + optimization sprint + monthly visibility monitoring.
Learn AI with This Course
What you’ll learn:
Learn how AI agents generate images and videos, evaluate visual quality, improve weak outputs, and build reliable workflows for mockups and multi-scene explainers.
Key topics:
Image and video prompt engineering
Visual evaluation with scores, judges, and rubrics
Autonomous agents for mockups and explainer videos
Until next week,
AI Radars