Anthropic's Claude AI model can now handle longer prompts
For Anthropic's API customers, the company's Claude Sonnet 4 AI model now has a one million token context window — meaning the AI can handle requests as long as 750,000 words, more than the entire Lord of the Rings trilogy, or 75,000 lines of code. That's roughly five times Claude's previous limit (200,000 tokens), and more than double the 400,000 token context window offered by OpenAI's GPT-5.
Long context will also be available for Claude Sonnet 4 through Anthropic's cloud partners, including on Amazon Bedrock and Google Cloud's Vertex AI.
Anthropic has built one of the largest enterprise businesses among AI model developers, largely by selling Claude to AI coding platforms such as Microsoft's GitHub Copilot, Windsurf, and Anysphere's Cursor. While Claude has become the model of choice among developers, GPT-5 may threaten Anthropic's dominance with its competitive pricing and strong coding performance. Anysphere CEO Michael Truell even helped OpenAI announce the launch of GPT-5, which is now the default AI model for new users in Cursor.
Anthropic's product lead for the Claude platform, Brad Abrams, told TechCrunch in an interview that he expects AI coding platforms to get a 'lot of benefit' from this update. When asked if GPT-5 put a dent in Claude's API usage, Abrams downplayed the concern, saying he's 'really happy with the API business and the way it's been growing.'
Whereas OpenAI generates most of its revenue from consumer subscriptions to ChatGPT, Anthropic's business centers around selling AI models to enterprises through an API. That's made AI coding platforms a key customer for Anthropic, and could be why the company is throwing in some new perks to attract users in the face of GPT-5.
Last week, Anthropic unveiled an updated version of its largest AI model, Claude Opus 4.1, which pushed the company's AI coding capabilities a bit further.
Generally speaking, AI models tend to perform better on all tasks when they have more context, but especially for software engineering problems. For example, if you ask an AI model to spin up a new feature for your app, it's likely to do a better job if it can see the entire project, rather than just a small section.
Abrams also told TechCrunch that Claude's large context window also helps it perform better at long agentic coding tasks, in which the AI model is autonomously working on a problem for minutes or hours. With a large context window, Claude can remember all its previous steps in long-horizon tasks.
But some companies have taken large context windows to an extreme, claiming their AI models can process massive prompts. Google offers a 2 million token context window for Gemini 2.5 Pro, and Meta offers a 10 million token context window for Llama 4 Scout.
Some studies suggest there's a limit to how large context windows can be, and AI models are not great at processing massive prompts. Abrams said that Anthropic's research team focused on increasing not just the context window for Claude, but the 'effective context window,' suggesting that its AI can understand most of the information it's given. However, he declined to reveal Anthropic's exact techniques.
When prompts to Claude Sonnet 4 are over 200,000 tokens, Anthropic will charge more to API users, at $6 per million input tokens and $22.50 per million output tokens (up from $3 per million input tokens and $15 per million output tokens).
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Gizmodo
a few seconds ago
- Gizmodo
Apple's AI-Enabled iPad on a Robot Arm May Interrupt Your Conversations (In a Positive Way)
Sure, Apple's AI-enhanced Siri rollout turned into a massive SNAFU that gave the Cupertino tech giant a rare black eye, but the company won't quit on Apple Intelligence just yet. The latest reports suggest Apple will try and sell consumers on at-home robots set to arrive in the next two years, along with a version of its AI assistant that sports a face and 'personality.' If the rumors are true, the company won't stuff AI into a ball-shaped bot that users hopefully won't trip over, but will put it on your tabletop in the form of a robot arm with a screen. That, and more Apple smart home products, are allegedly set to launch within the next two years. This robot, which my tech-addled brain can't stop referring to as 'Apple Intelligence on a stick,' will reportedly be able to interact with users on a more personal basis. We've heard about this supposed home robotics project before, but Bloomberg's star Apple leaker Mark Gurman wrote on Wednesday that Apple wants to give the bot the ability to interject itself into conversations as if it were an active participant. The iPad mini-sized screen at the end of an extending robot arm could show a cartoonish face that would be the most personified version of Siri yet. Apple has reportedly tested two versions of Siri's virtual identity. One is an animated version of the classic Finder logo on Mac. Another looks like a Memoji. I desperately hope Apple opts for the former. If you're trying to picture this AI in action, think less of the Samantha AI from the 2013 movie Her and imagine the Luxo Jr. mascot from Pixar Studios. This AI could be playful, awkward, and maybe even a little bashful. Apple showcased how this AI would work in a research paper it published back in January. The prototype AI framework, which Apple called ELEGNT, can follow commands and users' movements, but it may interact in more subtle ways that make it seem far more personable. Instead of outright telling you the weather forecast on command, the bot may look out the window, look back, and then say it's sunny with a chance of rain. As fun as it sounds, what matters more is if it's as functional as selecting an app on a screen. Home AI like Amazon's Alexa+ has rolled out slowly over the past few months. Early reviews of Amazon's at-home chatbot have noted that it can hold a conversation, but it's still not good enough to recognize specific context for your queries. Apple reportedly still has designs on humanoid robots, or at least something that will roll on wheels like TCL's AiMe or Samsung's Ballie. Those are still just prototypes, but we'll more likely see more Apple smart home products, like a doorbell camera and other home security devices. Gurman's slate of anonymous sources suggests Apple Intelligence on a stick may arrive in 2027, though we'll get a more subdued version that's closer to a smart speaker with a display before that. That device sounds like a HomePod version of an Echo Spot or Google Nest Hub, though this device should be wall-mountable. It won't have the conversational Siri, but it will have access to note-taking, video calls, and smart home controls through all the usual standby apps like Calendar, FaceTime, and Apple Music. All this could run in a new operating system that is currently codenamed Charismatic. None of this push for robotics will matter if Apple can't get its house in order, and that starts with Apple Intelligence and Siri. Apple is reportedly still planning to launch what it's dubbed internally as 'LLM Siri,' according to Gurman. That voice assistant may be available by spring next year, though we've been burned before on dates for Apple's previous AI rollouts. This new version may feature yet another visual redesign to help us forget about the first time Apple tried giving Siri an AI makeover. Whether Apple opts for in-house models or outside LLMs like Anthropic's Claude, we'll first have to deal with AI infecting our smartphones before we ever wrestle with any AI-enabled robotic arms.
TechCrunch
a few seconds ago
- TechCrunch
Waymo finally has a music experience worthy of its robotaxi
I'm riding in the back of a Waymo that's autonomously navigating the busy streets of San Francisco with relative ease thanks to 29 external cameras, six radar, and five lidar sensors all feeding into an AI model. For just 15 bucks, I get to experience what feels like a miracle of modern technology, and yet, there's a nagging thought I can't shake. The music sucks in here. Waymo's music streaming feature has felt like an aged barnacle attached to a futuristic shell. Until this week, passengers were limited to a few music stations that played lo-fi beats, smooth jazz, k-pop or other genres they may or may not care for. For those who wanted to listen to something more specific, they had to use another app from Waymo's parent company Alphabet. For iPhone users, that meant downloading the Google Assistant app and configuring it to connect to Spotify. At that point, you had to ask Google Assistant through written or verbal commands to stream certain songs, artists, or playlists on the Waymo. Even if you get to this point — at which you may be halfway to your destination and have listened to approximately three lo-fi beats — the service didn't work reliably. As I rode along in a robotaxi full of cutting edge technology, I was puzzled why Waymo had not figured out a simple way to stream music from my phone into the car's speakers — a breakthrough that automakers and audio manufacturers figured out a couple of decades ago. That's why I was pleasantly surprised this week to see that Waymo launched a Spotify integration allowing users to seamlessly link the music streaming and robotaxi hailing services. I immediately connected the services and hailed a Waymo to see how it would work. Listening to Spotify in a Waymo (Credit: Maxwell Zeff) Waymo's Spotify integration is nothing groundbreaking, but it adds to the user experience. It works seamlessly, which is roughly what I would expect when trying to play music on my car's speakers in 2025. But riding around in a Waymo, listening to my own playlist or picking up where I left off on a podcast, the backseat of the robotaxi feels more like my own space — which is increasingly the reason I opt for a Waymo. Techcrunch event Tech and VC heavyweights join the Disrupt 2025 agenda Netflix, ElevenLabs, Wayve, Sequoia Capital, Elad Gil — just a few of the heavy hitters joining the Disrupt 2025 agenda. They're here to deliver the insights that fuel startup growth and sharpen your edge. Don't miss the 20th anniversary of TechCrunch Disrupt, and a chance to learn from the top voices in tech — grab your ticket now and save up to $600+ before prices rise. Tech and VC heavyweights join the Disrupt 2025 agenda Netflix, ElevenLabs, Wayve, Sequoia Capital — just a few of the heavy hitters joining the Disrupt 2025 agenda. They're here to deliver the insights that fuel startup growth and sharpen your edge. Don't miss the 20th anniversary of TechCrunch Disrupt, and a chance to learn from the top voices in tech — grab your ticket now and save up to $675 before prices rise. San Francisco | REGISTER NOW To set it up, open the Waymo app and navigate to the 'Music' section, where you will notice a new offering that let's you connect to Spotify. From there, you can press a button and authorize Spotify to connect to Waymo, albeit, while giving the robotaxi provider some access to your listening information. Screenshot (Credit: Maxwell Zeff) Screenshot (Credit: Maxwell Zeff) I'm sure Apple Music users will soon want to stream their music and podcasts in Waymos as well. Waymo spokesperson Chris Bonelli told TechCrunch the company is always exploring new personalization options, but did not clarify when the company might add an Apple Music integration. Once my Waymo and Spotify accounts were linked, I just hailed a Waymo like I usually would, and got in. On the Waymo's touchscreen in the backseat, there's an option to select Spotify. I tapped it, and the podcast I was listening to on my headphones started playing from the exact spot I left off. Listening to TechCrunch's flagship podcast in a waymo (Credit: Maxwell Zeff) Image Credits:Maxwell Zeff You can turn 'autoplay' on or off in Waymo's Music settings, and upon entry, the robotaxi will automatically start playing whatever song or podcast you were listening to on Spotify. I liked having it on, but it does feel like it could get you into an odd situation if you're listening to an intense true crime podcast and then get into a Waymo with work colleagues. You can also use the Waymo's touchscreen to select from a variety of customized playlists that Spotify users will be familiar with, such as your 'Daylist' or other mixes. However, this selection doesn't seem to include albums, audiobooks, or podcasts that you've recently listened to. Choosing from different Spotify mixes (Credit; Maxwell Zeff) Luckily, your Spotify app now controls the music in the Waymo. You can simply select any song or playlist you want from your smartphone, and stream it throughout the vehicle like you would using Apple CarPlay or a bluetooth speaker. Controlling the Waymo's tunes with your phone (Credit: Maxwell Zeff) In the end, the Spotify integration made my robotaxi feel more personalized; I was even able to tweak the bass, subwoofer, and treble levels in the car's speakers. That personalization may not be the main attraction for first-time users, but it could keep them coming back — and a loyal customer base is exactly what Waymo needs.
Android Authority
a few seconds ago
- Android Authority
Gemini is bringing one of Gmail's handiest web features to your phone
C. Scott Brown / Android Authority TL;DR The Gemini 'Add to calendar' button in Gmail is now rolling out to Android and iOS. It automatically detects event details in emails and lets you add them to Google Calendar with one tap. The feature can handle multiple events in a single email, but only works in English for now. If you use Gmail on the web, you might've already spotted Gemini's 'Add to calendar' button offering to save you a few clicks. Now that same Gemini trick is coming to Gmail on Android and iOS. As Google explains in its latest Workspace Update, Gemini can now read through an email on your phone, pick out anything that looks like a calendar event, and pop up an 'Add to calendar' button right there in the message. Tap it, and a little panel opens in Gmail to confirm it's been saved to your Google Calendar. This saves you from switching apps or typing in details yourself. Don't want to miss the best from Android Authority? Set us as a preferred source in Google Search to support us and make sure you never miss our latest exclusive reports, expert analysis, and much more. It's a step up from what Gemini could already do on mobile. Since May, you've been able to ask it to create, edit, or delete events manually. This update adds some extra automation. Gemini does the spotting for you, and if there are multiple events in the same email, it'll offer to add them all in one go. There are a few limitations at this stage. It only works in English for now, and it won't show up for stuff Gmail already handles automatically, like restaurant or flight confirmations. Also, if you want to invite others, you'll still need to add them yourself. The rollout starts today for both Rapid Release and Scheduled Release Workspace domains, and Google says it could take up to 15 days before it reaches everyone. It's available for most Workspace plans, plus Google AI Pro and Ultra subscribers. Follow
