AI stories rising on Hacker News this week
The AI stories that gained the most upvotes over the last 7 days, counted while SteadyVane was watching.
The about 5× label next to an item means it climbed about 5 times as fast as the typical rising item SteadyVane tracks on Hacker News. An item without one wasn't clearly faster or couldn't be compared fairly.
These are SteadyVane's measurements, not fact-checked stories. Only items judged on-topic for AI tools are listed. Items found not to be in English are left out. Tap or click any name to see its numbers day by day. Updated every few hours.
The exact rules
Top movers per source over the last 7 days, ranked by how much the metric gained while SteadyVane watched. Only items the relevance reviewer scored on-topic for AI tools (6/10 or higher) are listed; an "about 5×" label means the item's gain per hour over its readings in these 7 days was about that many times the median gain per hour of the other items SteadyVane read on the same source in those days that climbed (each counted once, over readings at least an hour apart; a count that went down by more than 2% between two readings is left out of that median; such an item gets no label). A label is shown from 1.5× up, when at least eight other items climbed. YouTube videos whose views look paid for (almost no likes) are left out. So are items the language check finds are not in English. Rebuilt with every ingest cycle.
Also rose in the last 60 days
- We're releasing the weights of Mistral Large 4+1,647
- Mistral Large 4+1,575
- GPT-6 Sol and Luna+1,531
- Claude Opus 5.5+1,294
- Gemini 4 Argon+953
- Gemini 4 Argon+947
- AI-generated posters don’t have to be horrible+928
- Claude Haiku 5.5+898
- Kimi K3.1 Pro and K3.1 Pro Highspeed+815
- Claude Fable 5.2 and Mythos 5.2+765
- Goldman Sachs is pulling Claude Code from its bankers and traders+744
- OpenCode 2.0+733
- GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price+702
- Genie 4+702
- Tell HN: Gemini trains on chat history, even for subscribers (personal accounts)+693
- GPT‑6 and Intelligent UI for everyone+692
- AI is a tool and using it well takes skill+675
- Claude Fable 5.2 and Mythos 5.2+673
- The gap in shared understanding of LLM capability is widening+639
- Ask HN: What's your current AI agentic setup?+565
- GPT-6.1 Terra+560
- OpenAI seems to be silently re-enabling model training on Codex chats+528
- Dots: Always-on agents+498
- Survival LLM: AI assistant and emergency library that runs on your own computer+437
- OpenAI and Anthropic cut their LLM prices by up to 80% within a few months. Why?+416
- Show HN: Reladraw – A diagram language where you decide where to place things+404
- Agents don't need memory, they need documentation+370
- Claude Code regularly sends each background agent’s transcript to its model+366
- Claude Duet+327
- How the Claude Code team does code review+325
- Early progress on accelerating drug development with Claude+272
- Claude Opus 5.5 Intelligence, Performance and Price Analysis (Max)+258
- Getting the most out of Opus 5.5 in Claude and Claude Code+247
- Unreal Agent+245
- Talorys – A self-hosted personal AI agent on Cloudflare's free tier+235
- Show HN: Let your AI agents paint big arrows, boxes and text on your screen+234
- Show HN: BetterWispr – Free, open-source dictation for Mac+207
- MicroLLM Lab – Try 7 tiny LLM's in the browser+191
- OpenAI raises $11B at a $1.25T valuation+182
- Best LLM for every budget, updated daily+181
- ChatGPT Desktop forwards your SSH agent to remote machines+180
- Dayjob: Tax – evaluating AI agents on 50 professional tax assignments+179
- Claude Fable 5.2 scores 65 on Artificial Analysis, highest so far+154
- Basically no one is using AI agents for their DoorDash orders+133
- The Agent SDK+130
- Swedish government offices to ban ChatGPT and Claude+127
- Show HN: Made an open-source Lego AI generator+121
- Prompting Claude Opus 5.5+112
- Show HN: Hob - Open-source, on device alternative to Claude Cowork and ChatGPT+106
- Why are coding agents so dumb?+104
- Yes, Claude can do nine loops+100
- Microsoft abandons personal AI chatbot race with Copilot reboot+99
- Show HN: Atticus – Hold-to-talk voice assistant for macOS that runs on Claude -p+92
- Show HN: Omnara – Open-source alternative to Claude Managed Agents+90
- Ask HN: How do you manage your AI agents' context across clients and projects?+89
- GPT-6.1 Sol replaces GPT-6 Sol after just 7 days, with near-Astra intelligence+78
- If AI coding is lowering your code quality, you're not managing quality right+65
- Claude Duet+63
- Show HN: Pulse Local, a Mac budgeting app whose AI models run on your machine+60
- The Price of Intelligence Is Falling Rapidly+59
- "Just Ask Claude" Is a Rare and Valuable Skill+54
- Show HN: CUA-S1 – A System One Model for Computer Use+49
- Anthropic Subscriptions Offer 5x+ More Value Than OpenAI+42
- Claude Code now reads AGENTS.md if there is no Claude.md+34
- Show HN: Penmentor – Claude marks up your handwritten math like a teacher+32
- Jevmem – automatic project memory for Claude Code, built on Jev+12
- ZCode, the GLM coding agent, silently uploads your Git history+1




