Google Classroom gives teachers an AI feature for quiz questions

✨TodayOnAI’s Daily Drop

• Google Classroom gives teachers an AI feature for quiz questions

• OpenAI Launches GPT-4.1 Models Tailored for Coders and Agents

• OpenAI Retires GPT-4.5 API Access—Here’s What Comes Next

• 💬 Let’s Fix This Prompt

• 🧰 Today’s AI Toolbox Pick

Google

Google Classroom gives teachers an AI feature for quiz questions

🚀 TodayOnAI Insight: Google Classroom is expanding its Gemini-powered AI tools with a new feature that helps teachers auto-generate questions from custom text. This move deepens Google's push to streamline classroom prep for educators using AI.

🔍 Key Takeaways:

• The new tool generates questions based on user-provided text or uploaded files from Google Drive.

• Educators can customize output by grade level, question type (multiple choice or open-ended), quantity, and learning objectives.

• Questions can be exported directly into Google Docs or Google Forms for easy classroom use.

• The feature is exclusive to Workspace for Education users with the Gemini Education ($24/user) or Premium ($36/user) add-ons.

• Builds on Gemini’s prior capabilities like generating lesson ideas, vocabulary lists, and summarizing content.

💡 Why This Stands Out: By anchoring question generation to specific instructional content, Google is making AI more relevant to everyday classroom workflows. The granular filters—by skill, question type, and grade—suggest a thoughtful design for real pedagogical use, not just automation for its own sake. As Gemini evolves into a true teaching assistant, will it shift how educators plan and personalize instruction at scale?

OPEN AI

OpenAI Launches GPT-4.1 Models Tailored for Coders and Agents

🚀 TodayOnAI Insight: OpenAI has launched a new family of AI models—GPT-4.1, 4.1 Mini, and 4.1 Nano—designed to excel at coding, improve complex instruction following, and boost agent development. The models outperform GPT-4o and even surpass GPT-4.5 in key coding benchmarks, signaling OpenAI’s aggressive push to stay ahead of rivals like Google and Anthropic.

🔍 Key Takeaways:

• Model lineup: GPT-4.1 series includes three sizes optimized for code generation and software prototyping.

• Performance gains: GPT-4.1 scored 55% on SWE-Bench, exceeding previous OpenAI models, and is 40% faster than GPT-4o.

• Developer benefits: Offers 8x greater code context, improved bug fixing, and 80% lower input costs.

• Early feedback: Partners like Windsurf report 60% improvement over GPT-4o with fewer irrelevant file edits.

• Broader rollout: Available now via OpenAI’s API, with demo apps and livestream validation by OpenAI executives.

💡 Why This Stands Out: OpenAI’s coding-focused GPT-4.1 models sharpen its competitive edge in developer tooling and AI agent creation—both fast-growing sectors. As rivals flood the market with high-performing models, OpenAI is differentiating through speed, affordability, and developer-centric features. The shift hints at a broader trend: the next wave of AI dominance may come from tools that build and reason with software, not just text.

OPEN AI

OpenAI Retires GPT-4.5 API Access—Here’s What Comes Next

🚀 TodayOnAI Insight: OpenAI is retiring GPT-4.5 from its API just months after launch, citing high costs and strategic focus on future models. Developers have until July 14 to transition to GPT-4.1, the newly released alternative with similar performance at lower cost.

🔍 Key Takeaways:

• GPT-4.5 deprecation: OpenAI will end GPT-4.5 API access on July 14, just five months after its release.

• GPT-4.1 as successor: OpenAI launched GPT-4.1 as a drop-in replacement, claiming equal or better performance at a reduced price.

• Cost considerations: GPT-4.5 is expensive to operate—$75 per million input tokens, $150 per million output tokens—prompting its phase-out.

• Availability in ChatGPT: GPT-4.5 will remain accessible in ChatGPT’s research preview for paying users, despite API deprecation.

• Model background: Codenamed Orion, GPT-4.5 used more compute and data than any prior model but failed to meet "frontier" AI benchmarks.

💡 Why This Stands Out: OpenAI’s decision signals a strategic shift toward efficiency and scalability over sheer model size. GPT-4.5’s short-lived API run highlights the growing pressure to balance performance with cost—especially as AI workloads intensify. Is the era of ultra-large, ultra-expensive models already being rethought?

✨ See how a simple prompt upgrade can unlock better AI output.

🔹 The Original Prompt

At first glance, this prompt might seem okay. But it's too broad — and that limits the quality of AI-generated results. Let’s improve it using prompt engineering best practices.

✅ The Improved Prompt

💡 Why It's Better

• Specifies the target audience (beginner bloggers)

• Focuses on practical and relevant themes (SEO, content, monetization)

• Includes blog post formats (tutorials, listicles, etc.)

• Encourages variety and depth

🛠️ Learn how to adapt this prompt for SaaS, AI tools, dev teams & more →
Read the full PromptPilot breakdown

💡 Bonus Tool: Want to generate and master prompts instantly?
👉 Try PromptPilot by TodayOnAI (Free to use)

📰 More from the AI World

• Meta to start training its AI models on public content in the EU

• Hugging Face buys a humanoid robotics startup

• Nvidia says it plans to manufacture some AI chips in the US

• Meta’s vanilla Maverick AI model ranks below rivals on a popular chat benchmark

🧰 Today’s AI Toolbox Pick

• 🕵️‍♀️GoSearch (Apps Tool): Searches across work apps with multimodal-AI enterprise search.

• 🧠Predev (Developers Tool): Provides instant software planning and lightspeed development.

• 📬Chat Collect (Email Tool): Gathers emails from your GPT users to boost engagement.