Microsoft Launches OpenAI Competitors

Welcome to another edition of AI PlanetX.

OpenAI rolls out GPT-Realtime voice model; Microsoft introduces first in-house AI; Alibaba develops advanced chip for wider AI applications.

• Hottest AI News

• Top AI & SaaS Tools

• AI Tutorial: Create Product Images for Ecommerce

• Top AI & Tech News

• AI Art Spotlight

• Prompt of the Day: Turns Ideas Into Clear Instructions

• AI Video Tutorial

• Course of the Day: Agentic Knowledge Graph Construction

Hottest AI News

Microsoft

Microsoft Launches First In-House AI Models to Compete with OpenAI

Microsoft is making a bold move in the AI race by launching its first in-house AI models, signaling a move away from heavy reliance on OpenAI. The tech giant unveiled MAI-Voice-1 and MAI-1-preview, marking a fresh chapter in its complicated partnership with the ChatGPT maker.

Details:

• The MAI-Voice-1 speech model generates one minute of audio in less than a second on a single GPU. It already powers Copilot Daily’s AI news host and podcast features. Users can try it on Copilot Labs, customizing text and voice settings

• Microsoft trained MAI-1-preview with 15,000 Nvidia H100 GPUs for everyday queries. It will join Copilot’s assistant alongside OpenAI models. AI chief Mustafa Suleyman stressed a consumer-first approach, using Microsoft’s ad and telemetry data

• Microsoft plans to combine specialized models for different needs. This multi-model strategy aims to unlock more value and compete with GPT-5, DeepSeek, and others—while making Microsoft less reliant on outside providers

This launch represents Microsoft's determination to reduce dependency on external AI providers while building tech specifically optimized for its vast user base and data ecosystem.

The Future of AI in Marketing. Your Shortcut to Smarter, Faster Marketing.

Unlock a focused set of AI strategies built to streamline your work and maximize impact. This guide delivers the practical tactics and tools marketers need to start seeing results right away:

• 7 high-impact AI strategies to accelerate your marketing performance

• Practical use cases for content creation, lead gen, and personalization

• Expert insights into how top marketers are using AI today

• A framework to evaluate and implement AI tools efficiently

Stay ahead of the curve with these top strategies AI helped develop for marketers, built for real-world results.

Download the Free Report

OpenAI

OpenAI Launches GPT-Realtime, Its Most Affordable Voice AI Model Yet

OpenAI has launched GPT-Realtime, its most advanced speech-to-speech model, officially out of beta and ready for production. This marks a major leap in making voice AI more natural, affordable, and widely available.

Details:

• GPT-Realtime processes audio directly instead of the old transcribe-process-speak pipeline, cutting latency. It handles complex instructions, delivers expressive speech, switches languages mid-sentence, and adds two voices: Cedar and Marin

• The model also understands non-verbal cues like laughter, interprets images, and shifts tone. The updated Realtime API includes MCP (Model Context Protocol), like a USB port for AI data connections, key for e-commerce, travel, and customer service

• Costs dropped too: from 40 to 32 per million audio input tokens and 80 to 64 per million output tokens. Early adopters like Zillow report stronger reasoning and more natural speech, making property searches feel "like talking to a friend"

With thousands already building on the API and major firms reporting better user experiences, GPT-Realtime may be the push that brings advanced voice AI into mainstream business.

Top AI & SaaS Tools

• EasySite (Life-time Deal): AI full-stack software engineer that builds websites and apps through simple chat, featuring built-in databases, seamless mobile app conversion, and website-cloning capabilities ($100 off Tier 3 with code: AFChoudhury100)

• PixVerseV5: Image-and-text-to-video model with major improvements in motion, visuals, consistency, and prompt adherence [F-R-E-E until Sep 1]

• Google Vids: Create and refine short videos quickly using image-to-video, AI avatars, auto transcript editing, and multi-format export [F-R-E-E]

• Chance App: Visual agent that lets you point your camera at anything and get smart insights to better understand the world (available on Android and iOS) [F-R-E-E]

• Stitch by Google: Generate UI designs (Figma layouts) from text prompts — latest update adds canvas that shows the entire user flow at once [F-R-E-E]

Start learning AI in 2025

Keeping up with AI is hard – we get it!

That’s why over 1M professionals read Superhuman AI to stay ahead.

• Get daily AI news, tools, and tutorials

• Learn new AI skills you can use at work in 3 mins a day

• Become 10X more productive

Sign up and start mastering AI

AI Tutorial

How to Create Pro Product Images for Ecommerce

Gemini Flash 2.5 (nano-banana) makes pro-level product images simple — text-to-image, image-to-image edits, and multi-image fusion for ecommerce, here’s a guide.

---

1/ What this tool is

Gemini 2.5 Flash (aka nano-banana) is Google’s AI image generator and editor. It works for:

• Text to Image (generate from a prompt)

• Image to Image (edit, relight, swap)

• Multi-image fusion (blend multiple uploads)

2/ Where it’s most useful

Perfect for marketing and ecommerce:

• Amazon and Shopify hero images

• Lifestyle composites for ads and PDP galleries

• Quick colorway/SKU variations

• Bundled product arrangements

• UGC-style social shots

• Infographics with minimal text

3/ How to prep inputs

Have these ready before you start:

• 1–3 clean product photos (front, 45°, top)

• Licensed lifestyle backgrounds

• Brand HEX colors, fonts, and tone notes

• Platform specs (Amazon, IG, Pinterest, etc.)

• Compliance musts (nutrition facts, no false claims)

4/ Prompting basics

Use a structured format:

\[Subject] + \[Style] + \[Lighting/Camera] + \[Palette] + \[Composition] + \[Must keep/avoid]

Example:

“Single lipstick tube on marble counter — editorial product photo, soft daylight, neutral tones, top-down view, keep label intact, no text.”

For edits, start with scope words like “Edit only the background…” to limit changes.

5/ Advanced workflows

• Hero on white: “2000×2000, pure white #FFFFFF, soft contact shadow, centered at 90% frame”

• Lifestyle comp: “Place product on coffee table in living room, match lighting, scale realistically”

• Colorways: “Generate 3 bottle colors (sage, sand, navy) with cream cap, keep labels exact”

• Bundle: “Arrange 3 SKUs in shallow arc, unify shadows, export 2000×2000”

• UGC: “Smartphone selfie angle, natural window light, mild clutter in background”

6/ Troubleshooting tips

If outputs look off, refine with specifics:

• Labels blurry: “Preserve label exactly; do not redraw text.”

• Plastic sheen: “Keep natural texture; reduce gloss −30%.”

• Warped geometry: “Do not alter product silhouette.”

• Scene mismatch: “Match white balance to background; add soft shadow under base.”

Iterate with small, clear edits instead of long, complex prompts.

Top AI & Tech News

• Taco Bell is rethinking its rollout of AI voice assistants in drive-thrus after glitches and customer trolling exposed problems

• China's Alibaba has developed a more versatile AI chip (in testing) intended for a broader set of AI inference tasks

• GPT-5 outperforms licensed human experts by 25-30% and achieves SOTA results on the US medical licensing exam

• Krea launched a waitlist for a Realtime Video feature that lets users create and edit consistent videos using canvas painting, text, or live webcam feeds

AI Art Spotlight

Model: Midjourney V7

Prompt:

32-Bit Isometric view of and dark alley, with a police car with the sirens on, a detective near the car, raining --v 7

Prompt of the Day

Prompt That Turns Ideas Into Clear Instructions

This prompt transforms any product idea in your head into clear, actionable instructions that Cursor and Claude can execute immediately. Instead of struggling to turn your vision into code, it breaks down your concept into step-by-step tasks, technical requirements, and implementation plans.

Top AI Video Tutorial

Create Videos with Nano Banana, Runway, ElevenLabs (Realistic AI Video)

Complimentary AI Course of the Day

Agentic Knowledge Graph Construction

This course teaches how to design and implement a multi-agent system that identifies user goals, recommends which nodes and relationships to extract from structured and unstructured sources, and orchestrates specialized agents using Google’s ADK.