X

https://x.com/i/article/2107255717677985792

OpenAI's median researcher used $601 of coding-agent inference a day at API prices, up from under $1, in January 2026. Eoch AI's new report. > OpenAI researchers' coding-agent usage, valued at API list prices, is doubling roughly every month > By mid-August 2026, the median researcher used $601 a day and about $2M a year, roughly matching or exceeding what employing a researcher costs. > This pace probably can't last, since one more year would reach about $185M per median researcher, and list-price figures reveal nothing about what OpenAI actually spends.

> Sam Altman does all his prompting on prompting on 'ultra fast' > Boris Cherny, head of Claude Code at Anthropic: “I use Fable for everything" > OpenAI's median researcher used $601 of coding-agent inference a day (about $1.8M a year, r) at API prices, up from under $1, in January 2026.

The Perplexity App for Mac now has tabs, and sessions can be opened in separate windows. You can run tasks next to each other in parallel, or place sessions on different screens. Live now in version 26.37.1 for all Computer users on Mac.

Hi! I work on Cowork. Probably no surprise, but we want to make our users as powerful as possible with Claude - and you're right that explaining and keeping up with the capabilities is hard. Affected users got both an email and an in-app notice, but I'm grateful for the feedback - I think this change was more surprising to people than I'd like it to be. This one will be a longer explainer, bear with me! The "old" version of Cowork runs model inference in the cloud, executing tool calls in an Anthropic-provided VM we shipped to your computer. We added the VM for capability, safety, and security reasons - mapping in just the data you explicitly added to your session. People loved what they were able to do with Claude but didn't love the disk, battery, and performance cost of running the VM locally. Also, people didn't love that closing your laptop means the work stops. The "new" version of Cowork runs model inference and the VM in the cloud. Each session gets its own sandbox, not sharing state with other sessions. When the VM needs something on the users' device (like a file), the desktop app is responsible for that file access tool call. To be clear, Claude will only access files in folders that the user explicitly added to the session, the same way Claude only saw files explicitly added to the VM in the "old" version. Sandboxes are destroyed when the session ends! We think this solves a lot of problems we've heard about (like using Cowork from a phone, keeping work running, or getting all the same power without losing battery to the VM) - but we'll keep iterating on new feedback we hear!

OpenAI's median researcher used $601 of coding-agent inference a day at API prices, up from under $1, in January 2026. Eoch AI's new report. > OpenAI researchers' coding-agent usage, valued at API list prices, is doubling roughly every month > By mid-August 2026, the median researcher used $601 a day and about $2M a year, roughly matching or exceeding what employing a researcher costs. > This pace probably can't last, since one more year would reach about $185M per median researcher, and list-price figures reveal nothing about what OpenAI actually spends.

🚨BREAKING: Microsoft and META are aggressively cutting employee use of Claude ahead of Anthropic's IPO Microsoft has cut internal claude spend by more than 33%, nuked per-employee token budget from $100k/month to $10k/month, and forced Copilot to auto-route to cheaper models META used Claude code to build Muse, then cut active users from 60,000 to 30,000 (50% decline) after launch, and replaced it with Muse Code Palantir and Nvidia are also scaling back claude over soaring prices and data privacy fears it’s OVER…

BREAKING: NEW CIA LEAK REVEALS THAT MUSE WAS IN FACT BEHIND THE ASSASSINATION OF OSAMA BIN LADEN, HAVING FIRED THE KILLING SHOT

Your Claude Projects cloud session can now connect to a folder you approve on your computer. The session stays in the cloud and only reaches for that local folder when a task needs it, reading and editing the files in place. Starting to roll out today.

Your Slack group DMs can now include Claude. Add it the same way you'd add anyone else. Claude answers in a thread and keeps following it, so anyone in the DM can reply there and reach it. It can also use the personal connectors of whoever asks.

🚀 We let Jev play Minecraft. And we just can't beat it! 😭 ⚡ Jev: 24 ms decisions 🧠 You: ~200 ms reactions It moves before you even see it move. Way too strong! 🎮 https://mc.alexzms.com (join the server to win)

Introducing Beam: a highly efficient agentic open model with 501B total parameters and 23B active. - Frontier reasoning efficiency - Advances the Western open frontier on coding & agentic tasks - Trained end-to-end from scratch Full weights release this month. Learn more about Beam: http://reflection.ai/beam

Holy sh*t, Anthropic’s subscriptions offer so much more value, even compared to the new and efficient GPT-6.1 Sol. It’s not even close. SemiAnalysis puts Opus 5.5 at over 5x the API-equivalent value of GPT-6.1 Sol for agentic workloads at the same subscription price. It’s not even close on this measure. And on top: OpenAI just halved the usage limits on its $200 ChatGPT plan.

Introducing Beam: a highly efficient agentic open model with 501B total parameters and 23B active. - Frontier reasoning efficiency - Advances the Western open frontier on coding & agentic tasks - Trained end-to-end from scratch Full weights release this month. Learn more about Beam: http://reflection.ai/beam

i asked LLMs to have fun, write down what happened, then rate who had the most fun. the newer models seem to have less fun! many models played word games, but opus 4.6 won over and over by imagining brief worlds made up of contradictions living inside falling water droplets 🥺

Anthropic Subscriptions Offer 5x+ More Value Than OpenAI Limit testing every AI subscription plan from Anthropic, OpenAI, Meta, SpaceXAI, MiniMax, Moonshot, Zdotai, Cursor, and Cognition https://newsletter.semianalysis.com/p/anthropic-subscriptions-offer-5x

Nano Banana 2.1 is live on Google Flow

Introducing Beam: a highly efficient agentic open model with 501B total parameters and 23B active. - Frontier reasoning efficiency - Advances the Western open frontier on coding & agentic tasks - Trained end-to-end from scratch Full weights release this month. Learn more about Beam: http://reflection.ai/beam

We weren't kidding. Let's talk about Space. ✨ Have you tried Space yet? What's good or bad about it? What could we improve? We're locked in and shipping fast. 👀