Nvidia is in discussions to make a multi-billion-dollar equity investment in AI search startup Perplexity as part of a funding round valuing the company at over $30 billion. This valuation marks a more than 50% increase from Perplexity's previous round a year prior.

The financing round would be worth billions of dollars and would increase Perplexity’s valuation by more than 50% from its previous funding round a year ago, the Information report said. Reports last year had put Perplexity’s valuation at about $20 billion.

The potential investment comes as the two companies have developed stronger business relationships. Perplexity’s annualized revenue has grown to more than $750 million from less than $250 million at the start of the year, the Information reported.

The revenue growth has been driven in part by Perplexity Computer, an AI agent that professionals use to automate computer tasks, the Information reported.

The open-source, simple agreement that lets investors fund your startup now for shares of stock later. Created at Y Combinator by Carolynn Levy in 2013 and used to raise over $15B for YC portfolio companies.

How SAFEs work

Pick the fundraise amount. The valuation cap follows.

The biggest advantage of the post-money SAFE is that the amount of ownership sold is immediately transparent and calculable for both the founder and the investor. Start from how much you're targeting and how much ownership you're willing to sell — the valuation cap follows from the two.

Say you're targeting a $1M raise and 15% ownership sold. Your post-money valuation cap is $1M ÷ 15% = ~$6.7M, or "I'm targeting $1M at $6.7M post / $5.7M pre." Raise less on that cap and you sell less: $500k is ~7.5%, $800k is ~12%. You can also raise on more than one cap — $500k at a $5.5M cap (~9%) plus $500k at an $8.3M cap (~6%) gets you to the same ~15% sold.

Source: Y Combinator

Origin begins rolling out today in early beta on all paid plans. They're starting with the essentials, designed for agent scale: repos, pull requests, code browsing, and GitHub sync. Agent-native features ship soon.

Your GitHub repos can sit alongside the ones Cursor hosts. Connect GitHub to Cursor, pick your org, and you'll see the repos you can sync. Select one and Cursor pulls it in. You choose what gets synced and can disconnect a repo at any time. Anyone with read or write access to a synced repo can view it in Cursor too.

Synced repos update in real time. Browse, search, and pull from the copy in Origin. Pushes keep going to GitHub, which stays the source of truth for anything started there. Icons next to each repo name tell you which ones Cursor hosts and which came from GitHub.

Every repo has pull requests. Open one to see the timeline, commits, checks, and files changed. Review the diff, leave comments, and merge.

Pull requests on synced repos sync both ways: comment in Cursor and it posts to GitHub, react or reply on GitHub and it shows up in Cursor within seconds. Got a review assigned to you on GitHub? Review and merge it from Cursor.

Your code, PRs, and agents are now in the same place. Ask Cursor questions about code you're browsing. It can answer, make changes, update PRs, or push a branch.

Source: Cursor

OpenRouter is being acquired over five times the $1.3B valuation the AI routing startup raised at in May.

OpenRouter was found in 2023 and has around 8M users.

Future Claude models will generate text that contains a watermark. This is a way of determining the likelihood that Claude was involved in writing the text, and we, along with several other major AI providers, are implementing this change to comply with the EU AI Act.

As per Anthropic's statements:

• Claude is implementing watermarking to comply with the EU AI Act. Other major model developers have signed the same Code of Practice and will also be implementing watermarking;
• Their watermarking method doesn’t have any practical impact on the quality or content of Claude’s outputs;
• The difference between watermarked and un-watermarked text will not be distinguishable to readers;
• Nothing is added to the text and there are no hidden characters;
• Watermarking doesn’t require extra tokens, and will not be more expensive;
• Watermarks can’t be traced to a specific person, organization, or chat.

Source: Anthropic

Smallest AI gets support from Better Capital, Upsparks Capital, Schema Ventures, Tiny VC, DeVC, Mission Street Capital, and a group of angel investors who believe in the future of real-time voice AI.

Voice is becoming the primary interface between humans and machines, and smallest AI has powered over 1B minutes of real-time voice AI across 10 enterprises.

They claim to have passed the Turing test for Voice AI and raised $13M in Series A funding to bring the next generation of Voice to enterprises.

Source: Smallest AI