The AI landscape is heating up today with Snap making its big AI chatbot move, partnering with Perplexity in a $400M deal. Meanwhile, Google's expanding Gemini-powered navigation to India, showing that the AI race is going global at breakneck speed.

Snapchat just scored a $400 million payday from Perplexity, announced yesterday during Snap's Q3 earnings. Here's the twist: Perplexity is the one cutting the check—in cash and equity—for access to Snap's 943 million monthly users. The deal flips the traditional AI partnership model. Usually, platforms pay AI companies for their technology (like Microsoft paying for OpenAI integration). Now we're seeing AI companies pay platforms for distribution rights. For context: Perplexity has raised approximately $250M total. They're spending nearly 2x their entire funding just to access Snapchat users starting early 2026. Plot twist: Snap gets an instant AI search feature without building competitive infrastructure themselves, while Perplexity gets a direct pipeline to Gen Z eyeballs. It's like Snap just rented out prime real estate in their app and got paid $400M to do it. The signal: Distribution is becoming more valuable than model quality. Building the smartest LLM doesn't matter if no one uses it. Perplexity just bet $400M that user access beats technical superiority.

Today's Perplexity-Snap deal reveals something interesting about Perplexity's API strategy. While they haven't publicly released their API (unlike OpenAI or Anthropic), this partnership suggests they're maturing their infrastructure for enterprise integrations. Their model appears competitive with Claude and GPT-4o on web search capabilities, though with less transparent benchmarking. For builders: if you need real-time search in your AI app, watch Perplexity closely - they might open their API more widely soon. Until then, the Claude 3.5 Sonnet + web search combo (at $8/M tokens) remains your best bet for similar functionality with published benchmarks (94.0% on MMLU).

Today's $400M Perplexity-Snap deal signals a shift in AI investment strategy. Rather than just pouring money into model development, companies are now paying premium prices for distribution. This mirrors Microsoft's $13B OpenAI investment and Anthropic's $4B from Amazon - but with a twist. Instead of big tech buying AI startups, we're now seeing AI companies paying for access to users. The smart money suggests building proprietary AI is getting too expensive for all but the largest players, creating a new market for AI integration deals. Watch for more content platforms (TikTok? Reddit?) to announce similar partnerships in the coming weeks as the race for AI user adoption heats up.

A fresh GitHub project dropped today that creates structured guidelines for human-AI coding collaborations. The framework helps developers define clear motivations and requirements before sending prompts to AI coding agents - initially supporting Elixir/Phoenix but looking to expand. Think of it as creating a detailed treasure map for your AI pair programmer instead of vague directions.

Google's Gemini AI expansion to Indian Google Maps users today marks a major win for the company's global AI strategy. After successful deployment in the US, Indian users now get AI-enhanced navigation with contextual understanding of locations, traffic patterns, and safety concerns specific to Indian roads and neighborhoods. This rollout demonstrates Google's ability to adapt its AI for different markets and infrastructure challenges - a critical differentiator as competitors struggle to scale their AI solutions internationally.