... narrowed significantly over the past year. For prototyping or exploring a design, the $0.67 run is a good deal. For work requiring correctness and accuracy, Opus 4.7 remains ahead. Any experiences coding with open-weight models? [1]: MiniMax M2.7 vs. Claude Opus 4.6 Comment from fmerian(@fmerian): exactly - curious what's your preferred AI model when coding? see this thread Comment from Amit Raj(@amit_raj25): Is there any platform for renting GPU at reasonable price? Comment from Stan Kolotinskiy ... ... cost is really impressive! Comment from Stan Kolotinskiy(@sk_uxpin): @fmerian  hahaha, hard to disagree - especially for sensitive tasks Comment from Stan Kolotinskiy(@sk_uxpin): Interesting approach - I'm usually doing it the other way around: Sonnet/Opus

... Meow world, welcome back to The Breakpoint, a weekly thread on all things dev tools on Product Hunt. The latest Recent dev-first products launched on the site @Claude by Anthropic launched Sonnet 4.6, "the most capable Sonnet

... steps. (...) To handle longer running tasks, Composer 1.5 has the ability to self-summarize. This allows the model to continue exploring for a solution even when it runs out of available context. In a previous thread, Claude's (Opus, Sonnet) and Google's (Gemini) coding models were leading the way. [2] Curious if this new release would attract more developers. Any experiences with Composer 1.5? [1]: Introducing Composer 1.5 [2]: What's the best AI model for coding? Comment from ... ... Phalgun Guduthur(@phalgun_g): Tried it out for a whole day building something basic and I'm afraid I had to revert to Sonnet

... What would be the close 2nd? And if we look at the capabilities only, what would be the best model from your POV? Also, why not Opus 4.5!? Comment from fmerian(@fmerian): @derekattonkotsu  thoughts on the all-new Sonnet 4.6 ? Comment from fmerian(@fmerian): @atomer  any experiences with Composer 1? Comment from fmerian(@fmerian): good point - can't edit the poll! +1 @Claude by Anthropic Comment from fmerian(@fmerian): @iamasuperuser  oooooh good point. I forgot ... ... here Comment from fmerian(@fmerian): @dynamo  my favorite. to quote @rauchg , Opus is "on a different level, unreasonably good at @Next.js " - source Comment from fmerian(@fmerian): @Claude by Anthropic ftw! Comment from fmerian(@fmerian): @Claude by Anthropic family (Sonnet

... amongst this community of builders. I ll kick off with an experience that had us scratching our heads for months last year Our product, @Tonkotsu , runs a bunch of coding agents in parallel. They re built on top of Sonnet 4.5 and do coding tasks by repeatedly calling tools to read and write code. We review task failures daily and started noticing something strange: some sessions would devolve into an infinite loop of Sonnet

... generating a playable "Minecraft" replica demo cost less than 0.2 RMB using this model. Performance Benchmark: In major programming evaluation sets, Doubao-Seed-Code s overall performance is reported to be only second to the current top model, Claude Sonnet 4.5 , outperforming several other prominent domestic models. Massive Context: It features a native 256K context window , which is larger than Claude Sonnet

... shipping Let s crowd-source the best combos for 2025 . Who knows, maybe we ll spot the next trend before it blows up. Comment from mina(@minach): @kaadz  Very smart stack. Using Cursor/Windsurf first and falling back to VSCode/Qoder when limits hit is a great way to stay productive without breaking the budget. Free-tier optimization is an art Comment from mina(@minach): @ivan_apedo  Great combo .Fast UI first, then Claude/Cursor for the deeper logic. Smooth workflow ... ... building @raycast extensions, and not web apps. Comment from Maksim Ryndin(@maksim_ryndin): Front-end: mostly @Gemini , a bit of @Claude Code . In browser chats, prefer avoiding to give any access to the operating system. Backend: with @VS

... just launched AllPub.co and built pretty much the entire thing using Vibecoding with Claude Sonnet 3.5 . I'm a software architect with just some basic Python knowledge, so this was my first full web app build. Wanted to share my experience! What I Built AllPub.co is a tool that lets you write once in your favourite platform or Notion and automatically publish to Dev.to and Hashnode. I'm currently exploring more integrations like Medium, LinkedIn, and others based on what ... ... people actually want. No more copy-pasting and reformatting the same article three times. Why Claude Sonnet

... using both is the best way, Claude in Cursors Terminal brings me the code stability and I'm using it when deploys are failing or authentication , serious stuff. For new features im using cursors ai, mostly claude sonnet 4 max , Opus 4 is better but it costs far too much... Comment from Michael Mandeville(@mandeville): It's weird. Not a single person was using the terminal to write meaningful production code before Claude Code. No one I know is doing ... ... whatever keeps you productive, but mixing both could work well. Comment from Solal Zanovello(@solal_aegis): As a developer building a high-performance no_std Rust kernel (Aegis Omni-Core) , I ve found that the "Terminal vs

... results are mostly pretty good. A few things I ve come to love: Even though it s a CLI tool, it s pretty integrated into my IDE. When installing the CLI, it automatically installs the Jetbrains or VS Code extension, so you do get to see and manage the changes side by side in your IDE. The MCP functionality is pretty basic but it just works. With a good system prompt built over time, it knows when to run puppeteer ... ... Even considering the $200 20x Claude max plan (the current max plan has opus model usage run out in about an hour, and it is a better model) Comment from Kiran(@withkiran_): I am using the cursor with Claude Sonnet

... native hub to kill static PDF resumes. I wanted to create a standard where professional value is indexed via "Proof of Work" links (what we call the LCV standard). What tool? Mostly used Claude 3.5 Sonnet paired with Cursor and a bit of Gemini 1.5 Pro for architectural logic. Most difficult part? Implementing Privacy by Design . I didn't want a central user database. Getting the AI to help me build a "No-DB" architecture where all data is encoded ... ... first vibe code attempt was actually Anthos Adaptive Learning an AI-powered homeschool curriculum platform I'm building with my wife Nicole, a certified curriculum designer. Tool used: Claude (via Claude Code) as my primary coding agent inside of VSCode

... Anthropic just released two new models for Claude 4: Opus 4 and Sonnet 4 . Let's find out how good they are. What are you using it for? What s impressed you? What s broken? How does it compare to GPT-4, Gemini, and others? Drop your thoughts, screenshots, prompts, use cases below. Let s trade notes while it s hot Comment from Parth Ahir(@parth_ahir): @steveb  A multi-turn conversation is when ... ... back and forth, with the AI remembering context just like a real chat between people. (What we usually do with AI) Comment from Parth Ahir(@parth_ahir): Claude 4 s Opus and Sonnet

... Polite AI is excited to announce the integration of Anthropic s latest AI models, Claude Opus 4 and Claude Sonnet 4. These models are designed to enhance coding, reasoning, and agentic capabilities. Key Features: Claude Opus 4: Anthropic s most advanced model, capable of sustained performance on complex, long-running tasks and agent workflows. Claude Sonnet

... really bad coding assistant. I used it when the task was straight forward and simple (because it was free), but for anything that was complex or might touch more than one file, I fell back on Sonnet. I found Sonnet

... collaborative and productive environment. It offers participants the freedom to move around the meeting space (designed as various maps, some inspired by games and some based off of actual office blueprints). Nonilion is a distraction free google s meet alternative, with perplexity ai, multi participant screen share, interactive multi participant games, file sharing, custom characters and more. Nonilion is free to use for all the users and also offers a pro plan for meetings that required custom room names ... ... writers, video editors and front-end developers to focus and work in collaboration. Two or more groups talking over each other or sharing screens at the same time resulted in other groups to sit and just watch. As an alternative

... Codebase Bolt works well initially but can struggle as your codebase grows beyond roughly 100-200 prompts. Keeping your components short (under 300 lines each) helps manage complexity and Bolt's understanding of your code. Bolt uses the Claude Sonnet 3.7 model, which tends to be overly creative, suggesting unnecessarily complex solutions. Explicitly instruct Bolt, "I value simplicity and prefer straightforward implementations," to guide it towards simpler, practical solutions. Giving Bolt ambiguous prompts with multiple possible implementations often results ... ... fairly loaded topic. Analytics & Building Your Own Admin Tools Easily integrate Google Analytics by simply inputting your measurement ID into Bolt for basic user insights. Remember to ask it to track key events across your application. Alternatively

... Explain or prompt in a way that the AI models can understand better. It really is like working with a real intern :) Comment from Ajin Sunny(@ajinsunny): Nothing so far for me, if you are on the 3.7 Max Sonnet. It solved pretty much most with thinking. Comment from Eric Hubbell(@erichubbell): When updating an express app from ~v4 to ~v5 you no longer need try / catch statements for basic error handling. Express does this for you. So, I asked ... ... narrowed to the controller directory and finished in about 5 minutes. In sum, I don't like AI touching my code though I do enjoy the answers it gives me outside my project. The experience above was using Claude Sonnet

... Just a thought on Cursor and the fact that Anthropic rate limiting basically doesn't let me do anything right now. I've been developing with Cursor and Claude-3.5-Sonnet (recently 3.7-thinking) for last few months now. I've gotten to a point where I realized 1 hour of no-Cursor coding can be done in 10 minutes or less using Cursor agent. It's absolutely amazing, let me just say that. But because of the issues with ... ... Anthropic's endpoint, I was forced to try out gpt-4o, and other models. But nothing beats Claude-3.5(7)-Sonnet

... editor journey has been: 1. Tried out Github copilot in VS Code. Loved it. 2. Then tried out @Cursor after @bernatfortet kept talking about it. Immediately switch to Cursor. 3. Then tried out @Windsurf after @lagap discovered it. Immediately fell in love and switched. Anecdotally, everyone is talking about Cursor. And no one is really talking about Windsurf. On Product Hunt, Cursor has 845 Shoutouts. Windsurf has two? In my experience, Windsurf is the leader at the moment with ... ... like it so much more Comment from Gonzague Dambricourt(@gonzagued): @lagap  Totally agree Comment from Rashaan Thompson(@rashaan): Windsurf context awareness of project files and chat changed the game for me. I use it with the Claude sonnet