Meta releases Llama 4 models

Today’s Highlights:

📰 News: Meta releases new Llama 4 family of models + Microsoft upgrades Copilot into powerful AI companion

💰 Funding: Krea raises $47M for AI platform for creatives

⚡️ Top News Stories:

1. Meta has introduced Llama 4, its latest series of LLMs, featuring native multimodal capabilities that process and integrate various data types, including text, images, audio, and video.

• The initial models released are Llama 4 Scout and Llama 4 Maverick; Scout is designed to operate efficiently on a single Nvidia H100 GPU with a 10-million-token context window, while Maverick offers enhanced performance comparable to leading models like OpenAI's GPT-4o and Google's Gemini 2.0 Flash.

• Both models employ a "mixture of experts" architecture, activating only necessary components for specific tasks, thereby optimizing computational efficiency and resource utilization.

• These models show significant improvements in reasoning, code generation, and following instructions, with the 70B variant ranking among the top open-weight models across multiple academic and benchmark tests.

• Llama 4 models are multilingual, perform strongly across languages like English, Spanish, French, and German, and have a 128,000-token context window for more extended and nuanced conversations.

• Meta has also previewed Llama 4 Behemoth, an upcoming model with 288 billion active parameters, anticipated to be among the most powerful LLMs upon completion.

• These models are integrated into Meta's AI assistant, accessible through platforms such as WhatsApp, Messenger, Instagram Direct, and the Meta AI website

2. Microsoft is upgrading Copilot into a personal, context-aware AI companion with memory and autonomous agent capabilities, enabling it to proactively assist users across Windows, Office, Edge, and mobile devices—and even browse the internet and take web-based actions on their behalf.

3. Former OpenAI researcher Daniel Kokotajlo’s “AI 2027” report warns that superintelligent AI could arrive within two years, potentially triggering a technological explosion that dominates the global economy and demands urgent action on safety and governance.

4. OpenAI has reversed its earlier decision and will now release both the o3 and o4-mini reasoning models ahead of GPT-5, which has been delayed to allow for deeper integration and infrastructure scaling.

5. Midjourney has launched V7, its first new image model in a year, featuring default personalization, improved coherence with prompts and object rendering, and a new Draft Mode for fast, low-cost image generation.

6. Google is expanding its AI Mode in Search with a new multimodal feature that lets users ask complex questions about images they upload or capture, using Google Lens to interpret entire scenes—including object relationships, materials, and spatial arrangements—and deliver AI-generated responses.

7. Amazon’s Nova Reel 1.1 can now generate up to two-minute-long videos composed of multiple stylistically consistent shots, including a new manual mode that allows image-guided control — all accessible through AWS services like Bedrock.

8. Shopify CEO Tobi Lütke has mandated that teams justify why AI can’t do a job before requesting more staff, aligning the company’s resource decisions with a growing industry push toward AI-driven efficiency amid broader concerns about job displacement.

9. GitHub Copilot is introducing new usage caps and a premium pricing tier for more advanced AI models like GPT-4.5 and Claude 3.7, signaling rising compute costs while underlining Copilot’s rapid growth and profitability within Microsoft.

💰 Top Funding News:

1. OpenAI is in discussions to acquire io Products — an AI hardware startup co-founded by CEO Sam Altman and famed Apple designer Jony Ive — for $500M, as the team explores smart home devices aimed at being less disruptive than smartphones.

2. Krea, which offers a unified AI platform for creatives to generate and edit visual content across multiple models, raised a $47M Series B, led by Bain Capital Ventures with participation from Andreessen Horowitz and Abstract Ventures.

3. Moonvalley, which develops AI-powered video generation tools with a focus on customization and legal safeguards, raised a $43M round, led by unnamed investors.

That's all for today's email! If you want more please follow us at the social channels linked below, or check out our website!

Share our newsletter: If you like our work please share/forward this email with your friends, colleagues, and family. It's the best way to support us!

If this email was forwarded to you please sign up here to continue receiving them.

Want your content, product, jobs, or event featured in our newsletter? Reply to this email with the details, and our team will reach out to you.

Do you use AI for work? Tell us how, and you could be featured in our newsletter!

Check out our website for more resources, including a list of AI investors, products, events, and twitter follows.

For an archive of all our posts, click here.

We'd love to hear from you! You can always leave us comments or feedback by replying to this email!

Powered by AI. Curated and edited by Humans.