Perplexity Launches Web Browser

Good morning. It’s Wednesday, February twenty sixth.

On at the present time in tech history: In 1991, Tim Berners-Lee, a British computer scientist, introduced the world to “WorldWideWeb,” the primary web browser and a WYSIWYG (What You See Is What You Get) HTML editor.

Claude Sonnet 3.7
Perplexity Web Browser
AI Audio Generation for Video
OpenAI drops Deep Research access to Plus users
DeepSeek Rushes For Recent AI Model
4 Recent AI Tools
Latest AI Research Papers

You read. We listen. Tell us what you think that by replying to this email.

Atla launches Selene 1, a state-of-the-art LLM Judge

Atla releases Selene 1, an LLM trained to judge AI responses for quality, correctness, and more. As a substitute of generating answers, Selene scores and critiques them, and does so more effectively than some other model in the marketplace.

Selene beats frontier models from leading labs–including OpenAI’s o-series, Anthropic’s Claude 3.5 Sonnet, DeepSeek’s R1–across 11 commonly used benchmarks for evaluators.

Selene might be used for a spread of evaluation tasks, from detecting hallucinations to checking correctness in specific domains. Its instruction-following capabilities also enable users to create and run evals with custom evaluation criteria.

Selene is obtainable via API/SDK. Atla also released an Alignment Platform, which helps developers generate, test, and refine custom evaluation metrics for specific use cases.

Today’s trending AI news stories

Claude 3.7 Sonnet Introduces ‘Pondering Mode’ to Balance Speed and Depth in AI Responses

Anthropic is raising the stakes in enterprise AI with Claude 3.7 Sonnet, a model that lets users fine-tune its depth of reasoning on demand. The ‘considering mode’ toggle shifts between quick responses and deep evaluation, difficult OpenAI’s modular approach and DeepSeek’s cost-cutting strategy. Benchmarks put Claude 3.7 at 78.2% accuracy on graduate-level reasoning tasks, outpacing DeepSeek-R1 in instruction-following. Slightly than chasing raw speed, Anthropic is betting on AI that adapts to complexity without switching models.

Introducing Claude 3.7 Sonnet: our most intelligent model to this point. It is a hybrid reasoning model, producing near-instant responses or prolonged, step-by-step considering.

One model, two ways to think.

We’re also releasing an agentic coding tool: Claude Code.

— Anthropic (@AnthropicAI)
6:30 PM • FEB 24, 2025

Alongside Claude Code—a command-line AI developer assistant with built-in human oversight—the corporate is positioning itself as the choice to fragmented AI stacks. If enterprises buy in, this might mark a shift away from model juggling toward AI that really thinks before it speaks. Read more.

Perplexity teases web browser ‘Comet’

Comet is touted to be an agentic web browser. (Image: Perplexity)

Perplexity AI is taking its search ambitions up a notch with Comet, an AI-powered web browser. While details remain scarce, Perplexity claims Comet will “reinvent” browsing, very like it did with search. This move follows the corporate’s streak of rapid expansions—an AI assistant for Android, a deep research tool, and an API—each reinforcing its bid to own the AI-first web experience.

Despite its growth, Perplexity faces mounting legal scrutiny. As Perplexity scales its monetization efforts including promoting and enterprise offerings, Comet’s success may hinge on whether the corporate can navigate each technical innovation and legal headwinds without losing momentum. Read more.

Luma AI Adds Free Audio Generation to DreamMachine Video Tool

Luma AI has added audio generation to its DreamMachine video tool, allowing users to create soundtracks for his or her AI-generated videos. The brand new “Audio” feature enables automatic sound generation with a single click or customized audio using text prompts.

Video to Audio is now here in #DreamMachine. To generate sound to your video generations, just select the brand new “Audio” button. Create with a single click or describe with prompts for more customized direction. Audio is obtainable now in beta without spending a dime to all users.

– before (@melabairs)
4:34 PM • FEB 24, 2025

Currently in beta, the feature is obtainable without spending a dime to all users. This update expands DreamMachine’s capabilities, integrating AI-driven visuals and audio inside a single workflow. Read more.

OpenAI drops Deep Research access to Plus users, heating up AI agent wars with DeepSeek and Claude

OpenAI has expanded access to Deep Research, its AI-powered evaluation engine, bringing ChatGPT Plus, Team, Education, and Enterprise users into the fold. Running on a specialized o3 variant, it autonomously pulls, structures, and synthesizes insights across text, images, and PDFs—delivering intelligence that mimics human evaluation.

It’s a calculated move. DeepSeek’s open-source R1 is shaking up the AI market, undercutting OpenAI’s paywalled grip. Perplexity is running R1 for a fraction of the associated fee, while Anthropic’s Claude 3.7 Sonnet leans into transparency with explicit reasoning trails.

Subscription tiers, nevertheless, remain firmly in place. Plus users get 10 queries a month, while Pro subscribers—paying ten times more—unlock 120. Read more.

DeepSeek rushes to launch recent AI model as China goes all in

Fresh off the market-shaking debut of R1, DeepSeek is now accelerating the launch of its R2 model, originally set for May. R2 is predicted to reinforce code generation and expand multilingual reasoning, further strengthening DeepSeek’s competitive position.

Once viewed with skepticism by Beijing over its massive chip acquisitions, DeepSeek now enjoys quiet state backing, with government-linked enterprises integrating its models at scale. Nvidia recently teamed up with DeepSeek to optimize R1 for Blackwell, driving a 25x revenue increase and 20x lower inference costs. Developers can now access an FP4-optimized DeepSeek checkpoint on Hugging Face.

The corporate has also resumed API access after a three-week pause resulting from capability constraints. Developers can now top up credits, though server strain persists during peak hours. Its domestic rivals are accelerating development—Alibaba launched a preview of its own reasoning model, QwQ-Max, on the identical day DeepSeek reopened API access. Read more.