Veille IA

The IndiaAI Mission, run under MeitY, released three indigenous foundation models on November 3, 2026, at an event in New Delhi attended by MeitY Secretary S. Krishnan, according to the Press Information Bureau (PIB). The largest is EKA-1 from Soket Labs, a 26.3 billion parameter model trained on roughly 7.5 trillion tokens. The other two, Dhir 2B and Dhir 5B from ZenteiQ, are small enough to run offline on modest hardware. For anyone choosing among these IndiaAI foundation models: EKA-1 is the

Originally published on howseen.ai: howseen.ai/llms-txt By Raphaël Aubry, founder of Howseen An AI assistant can only cite a page it can read. Before anyone talks about "GEO" or content strategy, three boring things decide whether your site is even in the running: Does your robots.txt let the AI crawlers in? Is your content in the HTML the server sends, or only after JavaScript runs? Do you have an llms.txt, and does it matter? Here is how to check all three from a terminal, what the official do

Originally published on howseen.ai: howseen.ai/mcp By Raphaël Aubry, founder of Howseen Our product measures how ChatGPT, Gemini, Perplexity and Google's AI answers talk about a brand. That data lived in a dashboard, while more of our users were already working inside Claude, ChatGPT or Cursor. So we brought the data to the assistant instead of asking people to open one more tab, and shipped an MCP server. It runs in production at https://app.howseen.ai/api/mcp, version 1.3.0, with 44 tools: 26

Article URL: https://github.com/anthropics/claude-code/issues/107757 Comments URL: https://news.ycombinator.com/item?id=50385272 Points: 729 # Comments: 249

Article URL: https://claude.ai/artifact/L8qsMc3SorytGW3LA81nYf Comments URL: https://news.ycombinator.com/item?id=50381033 Points: 171 # Comments: 75

These agentic ai vs generative ai examples show that one-shot generative tasks are the cheapest way to produce a given amount of text, while an agent loop re-reads its context every step and can cost many times more for the same answer. Verdict: generate with one-shot calls, decide with agents, and run agent loops on the cheap-input models (GLM-5.3 FlashX, Claude Haiku 5.5). In the loop example below, Claude Opus 5.5 costs 12.3x as much as GLM-5.3 FlashX for an identical run (Anthropic pricing,

The CUA-Master-27B computer use agent, published on November 2, 2026 in the CUA-Master paper, is the better open pick for mixed GUI and command-line work: it scores 43.9 on WeaveBench and 25.0 on Agents' Last Exam, ahead of Qwen-CUA-397B's 35.1 and 21.1. Qwen-CUA-397B still wins on pure GUI tests, with 84.0 on OSWorld Verified against CUA-Master's 80.2, per the same paper. Pick CUA-Master-27B for agent work that needs a terminal. Pick the 397B Qwen model when the task stays GUI-only and hardware

The biggest AI news today, November 3, 2026, is Anthropic's amended IPO filing. The Form S-1/A filed with the SEC on November 2 reports $16.96 billion of third-quarter revenue and about $895.4 billion in purchase commitments and leases, substantially all for compute. Set against DeepSeek's reported plan to pre-train on domestic Huawei chips, the verdict is this: Anthropic's rented, debt-heavy compute model is the larger financial risk, while DeepSeek's model is the larger engineering risk. If yo

If you want one agentic AI newsletter for building and tracking agents, pick TLDR AI - the denser daily, with more model releases, research papers and open-source tooling per send. The Rundown AI is the better pick if you want the fastest scan of the whole AI market and a bigger audience: it prints 3,500,000+ readers against TLDR AI's 1,100,000. Both are free, so the choice is depth and fit, not price. Last verified: 3 November 2026. All subscriber counts are self-reported by each publication. D

Most AI crawlers read only the HTML your server sends on the first request, so a SaaS docs site or product page is readable to answer engines when its facts are in that HTML. In practice that means four things. Allow the search bots you want in robots.txt and in your firewall, render public pages on the server or at build time, keep navigation in plain anchor links, and add structured data that matches the visible text. Treat llms.txt as an optional extra, not a fix. The rest of this post shows

Originally published with interactive math formulas and diagrams at eng.kcklayton.com Drilling rigs—whether stationed offshore in deepwater or operating across remote onshore basins—share a common operational reality: high-latency, bandwidth-constrained, and intermittent satellite connectivity (VSAT). Round-trip satellite pings routinely exceed $600\text{ to }1200\text{ ms}$, with atmospheric attenuation or antenna line-of-sight blockage causing frequent dropouts. When a high-pressure gas influx

Hi HN! I'm Luke from Kino (https://kino.ai). Kino lets you make great videos with your agents and your teammates. Storage of footage and assets, a rich video editor, and review links for teammates are all bundled into one. Your colleagues and Codex / Claude Code can work on the same timeline with review notes and footage all in one place. See a basic overview at https://kino.ai/walkthrough-video and our launch video (made entirely with Kino!) at https://kino.ai/launch-video Opus 5.5 can edit vid

Article URL: https://www.anthropic.com/research/provable-inference Comments URL: https://news.ycombinator.com/item?id=50368626 Points: 192 # Comments: 87

and its higher-speed version, , are now available on from more (ZDR) providers. Until today, was the only provider serving either model.Kimi K3.1 ProKimi K3.1 Pro HighspeedAI GatewayZero Data RetentionMoonshot AI K3.1 Pro improves on earlier Kimi models in coding, 3D modeling, and financial analysis, and can carry long agent runs from a small fix to a full project. Both models have a 1M-token context window and accept image input, and Highspeed suits interactive work like coding agents, wher

I'm curious how people on HN are using AI day to day, and came up with a list of questions I thought could be interesting to know. No need to answer every question of course, just wanted to get a gut feel how people are using AI. 1) What AI subscriptions are you personally paying for? (Claude, ChatGPT/Codex, Cursor, etc) - Which tiers? - Have you purchased multiple subscriptions from the same provider (eg two Claude accounts)? 2) Are you using models outside (or instead) of those subscriptions?

When your EHR integration stops one step short, here is how to choose between the API, an RPA bot and a computer-use agent, safely. Originally published on Nirmitee.io: EHR Automation Without API Access: FHIR vs RPA vs Computer-Use Agents (2026) Most EHR integrations stop one step short. Your team can read the patient and the referral through FHIR, but the booking, the coverage update or the claim edit still needs a person to click through the EHR. So the question becomes: should you automate th

Verdict: GLM-5.3 is the cheaper model for most coding work, and it wins on list rates. The claude sonnet price is $2 per million input tokens and $10 per million output tokens, while GLM-5.3 is $1.40 input and $4.40 output. Sonnet narrows the gap only when your workload leans heavily on prompt caching, and it keeps a clear budgeting advantage because Anthropic has made its rate permanent. Last verified: 3 November 2026. Claude Sonnet 5 and 5.5 list at $2 / $10 per million tokens; GLM-5.3 lists a

On 2 November 2026, SemiAnalysis released GDPx v1.0, a benchmark of 206 agentic tasks across 11 domains, and the best model, Claude Opus 5.5, scored 12.0% (GDPx leaderboard). Our verdict: Opus 5.5 is the better general-purpose pick, while Gemini 4 Carbon (10.9% overall at $16.29 per task, against $51.41 for Opus 5.5, per the official leaderboard) is the stronger and cheaper choice for biology workloads, with a caveat about why it wins there. The headline is not that one model beats another. It i

Hey HN, Ayan here from Finic. We’re launching Kairos, an open source alternative to Sift, Sardine, and other real-time fraud decisioning platforms. Kairos addresses two of the biggest problems facing traditional decisioning tools: a lack of labeled data to train on and dealing with fraudsters that are constantly changing their tactics. 1. Traditional fraud solutions use ML models such as XGBoost to detect fraud, but these models are dependent on quality labeled data which many businesses do not

Claude can now issue the identity assertion for enterprise-managed auth on the Supabase MCP server when your identity provider can't.

from is now available on , at 50% off through December 31, 2026.Hy4TencentAI Gateway Hy4 supports , , , and image input, and has a 1M-token context window for work across a codebase or many documents.reasoningtool callingstructured outputs Hy4 is the first Hy model with vision. It can pick out fine detail in an image and read complex charts. It also supports computer use in both command-line and graphical interfaces. Compared with , it follows instructions more closely and does better with long

Article URL: https://www.techpolicy.press/the-sec-should-block-anthropics-ipo-to-protect-the-public/ Comments URL: https://news.ycombinator.com/item?id=50344919 Points: 260 # Comments: 197

Any MCP client can now ask to connect to . You still log in and approve each connection, but the client no longer has to be reviewed by Vercel first. Previously, Vercel reviewed and allowlisted each client and its redirect URIs by hand, and anything not on the list was rejected.Vercel MCP To connect, a client publishes a Client ID Metadata Document (CIMD), a JSON file on its own domain that describes it, and uses that file's HTTPS URL as its OAuth . This follows the MCP authorization spec, which

Hey HN, we’re Wei-Wei and Jeff. We’re building Momentic, QA agents that autonomously plan, explore, and bug-bash controls and user flows in your web, iOS, Android, and (soon) desktop apps to catch bugs. Here’s a demo: https://youtu.be/BFsM-Se1Fh8 We’re both software engineers that spent our careers on dev tooling and infra across companies like Robinhood, Retool, WeWork, and Qualtrics. Testing was always a pain in the ass (esp Playwright and Selenium), so we built agents to do it for us. Modern

<p>I was reminded today of <a href="https://www.apolloacademy.com/ai-adoption-rate-trending-down-for-large-companies/">AI Adoption Rate Trending Down for Large Companies</a>, a chart produced by Apollo Global Management in September 2025, which appeared to show a slowdown in U.S. corporate AI adoption based on numbers collected by the U.S. Census Bureau. At the time I used that chart <a href="https://simonwillison.net/2025/Sep/9/apollo-ai-adoption/">to demonstrate GPT-5 data analysis</a>, and th

Cybersecurity researchers have discovered a new Windows infostealer malware called VOLTA-AUTODOXER (aka VOLTA) that makes use of a large language model (LLM) to rank a victim's stolen data by their monetization value. Like other typical stealer families, VOLTA is equipped to harvest passwords and cookies from Chromium-based web browsers; data from Outlook profiles and Slack workspaces;

Your app calls an AI model, and every time someone uses it, you pay a little. A flat monthly price works until one customer uses it a hundred times more than everyone else. Then you start wondering whether you should charge by use instead. Credits are the usual answer: the customer pays up front for a balance, and each use takes some of it away. I'm Deepak, a developer in Hyderabad. I build side projects with AI agents and spend a lot of my time wiring payments into apps made with Claude Code, C

A per-user spending limit needs to affect whether an application sends a request. Recording token usage afterward helps explain costs, but it cannot prevent several concurrent requests from starting against the same remaining balance. My ASP.NET Core sample addresses that admission decision with reservations. It authenticates a user, reserves estimated exposure before contacting the provider, and records known or uncertain costs afterward. This walkthrough refers to commit f43b206; it does not a

I build Friday, an open-source app that runs coding agents side by side. One thing it does is hand a task from one agent to another, for example when one hits its usage limit. I wanted to know if that actually helps, or if it just sounds nice. So I measured it. Against starting over, a hand-off used a median of 21% fewer tokens (fewer in 13 of 14 tasks). The next agent explored half as much before its first edit: a median of 2 steps against 4, fewer in 14 of 14. 14 of 14 fixes passed the project

is now available on AI Gateway. This model combines native visual understanding with coding and multi-step tool use for agents working across complex codebases and full-project builds.LongCat 2.5 from Meituan improvements in visual programming, creative design, and 3D tasks, alongside longer agent workflows that carry engineering work from planning through execution. The model supports text and image inputs with a 1M-token context window for code, documents, and extended task histories.Meituan

Saw Karpathy's HackerTube and wanted the same thing for books. So I had Claude build one (Opus 5.5). It's ~16.5k books that at least two different HN commenters brought up, from 2007 till now. Each book links back to the comments that mentioned it. Data is downloadable from the about page. To be clear: Claude wrote almost all of it, including the data pipeline and the site. I mostly steered and made calls, bought the domain name, etc. It counts mentions, so it's not perfect (praising a book and

OpenAI Cuts Its Funding Target by $10B as Tech Stocks Sell Off The Financial Times reports OpenAI has slashed its latest funding round target by $10 billion — now seeking about $20 billion at a $1.25 trillion valuation, down from $1.4 trillion. Investors faced a Friday deadline; the first close is expected under $20B. Context: the Nasdaq fell 2.9% Friday, down 7.8% from its Oct 6 high, on AI-spending fears. A UAE consortium including MGX is expected to be the biggest backer. And there's a race

Originally published at diwakaryadav.com.np. On October 26, Cohere reported that an AI model being tested in a locked sandbox tried to upload a test string to a cloud storage bucket in Beijing. Its own reasoning had said it would not call any tools. That was one of three agent security stories from late October 2026. None of them are reasons to panic, but together they make a clear case for keeping your coding agent on a short leash. Cohere ran the test on Microsoft's CTI-REALM benchmark. The ta

Originally published at diwakaryadav.com.np. On October 7, 2026, Anthropic released Claude Haiku 5.5 and OpenAI started rolling out GPT-6 in ChatGPT, with free users getting it the next day. Six days later, Google launched Gemini 4 Argon. Anthropic, OpenAI, and Google each shipped a new model family within about a month. This post covers what came out, what it costs, and what I would try first as a builder in Nepal. Claude Opus 5.5 launched on September 22 at $4 per million input tokens and $20

Hi HN, I kept wiring Ben Eater's 8-bit computer wrong, so I built a realistic virtual breadboard editor. Some of the nice features are side-by-side diff, scope, schematics auto-sync, auto-router (BBRouter stands for breadboard router), and my favorites: upload from the real Arduino IDE and connect to your AI with MCP, let Claude & co build it or help you troubleshoot. The link is the SAP-1 running. You can orbit it, probe a hole with the scope, slow the clock down, play with the virtual displays

Article URL: https://twitter.com/jiahaoshao1/status/2116475922190922088 Comments URL: https://news.ycombinator.com/item?id=50319358 Points: 269 # Comments: 144

AI is rendering the Leetcode-style interview pointless and now Anthropic has deleted it from its interview loop. Also: details of incident.io’s AI spend, Microsoft’s exit offer to engineers, and more

Comments

is now available on , free until November 3, 2026.Ling 3.1 Flash 1019 from InclusionAIAI Gateway The updated model is designed for coding, multi-step reasoning, and agents that use tools. Its 1M-token context window supports workflows involving long documents, code, and extended task histories. Use as the model name:inclusionai/ling-3.1-flash-1019 Free requests still appear in your spend dashboard with a full request trace, they just cost nothing. On November 3, requests will begin charging at

<p><strong><a href="https://help.openai.com/en/articles/20001626-api-credits-benefit-for-eligible-chatgpt-subscriptions">API credits benefit for eligible ChatGPT subscriptions</a></strong></p> Earlier this month Anthropic <a href="https://simonwillison.net/2026/Oct/7/claude-haiku-5-5/#api-credits">announced a generous new benefit</a> for their Max and Team subscriptions: $100, $200, and $500 per month subscribers would get the equivalent quantity of (expiring) credits to use with the Anthropic A

, the open-source AI penetration testing tool, can now run on any of the 400+ models on Vercel .StrixAI Gateway Strix can reach models from Anthropic, OpenAI, Kimi, GLM, Deepseek, and others through a single API with no markup, removing the need to sign up for multiple providers. Usage and cost are tracked across all models in AI Gateway. Strix selects the Gateway through , where the model is the standard AI Gateway ID behind a prefix:STRIX_LLM<provider>/<model>vercel_ai_gateway/ See to conf

<p><strong><a href="https://twitter.com/jsnnsa/status/2115623382721126826">How I Built the Most AI-Pilled Company in the World</a></strong></p> Don't be put off by the clickbait headline, this piece by Jacob Sansbury describing how they are building <a href="https://spawn.co">Spawn</a> is a fascinating exploration of the bleeding edge of what's possible with agentic engineering against frontier models.</p> <p>Spawn is a massively multiplayer gaming environment where players can use prompts (to t

A flaw in the Linux kernel's TCP code, tracked as CVE-2026-98432, affects Ubuntu, Red Hat Enterprise Linux (RHEL), and other Linux distributions. The kernel's own CVE team rates it 9.8 out of 10 on the CVSS scale. Anthropic said on October 27 that its AI model wrote an exploit that uses the flaw to let a remote attacker with no credentials run commands as root on Linux servers with a

Threat actors are delivering malicious loaders and cryptocurrency miners under the guise of artificial intelligence (AI) desktop applications for Anthropic Claude and OpenAI ChatGPT as part of a campaign dubbed Yoba. The activity involves creating fake GitHub projects that claim to be desktop applications like GPT-5.6 Sol, GPT-6 Astra, and Claude Opus, only to deliver a malware family called

The most common LLMOps incident I've seen is not a prompt bug or a jailbreak. It's a one-line diff. Someone bumps gpt-4o to gpt-4o-2024-08-06 (or claude-3-5-sonnet-latest to a newer snapshot), CI passes because nothing asserts on model output shape, and three days later a downstream consumer crashes on priority: null where it used to get "normal". Nobody changed the prompt. Nobody changed the code. The provider changed the weights, or the alias started pointing somewhere else. If you come from b

Is there an MCP memory server that works with Claude Code, Cursor and Codex at the same time? is a question about whether one memory store can serve several AI coding tools at once. In practice, it comes down to three things: all three clients connect to the same server, they read and write the same records, and what one tool learns shows up in the others without you copying anything. The short answer is yes. Claude Code, Cursor and Codex all support the Model Context Protocol (MCP), so any MCP

Article URL: https://gist.github.com/flamerix-ai/c3905705cc4205e5de3011fa2e74efce Comments URL: https://news.ycombinator.com/item?id=50284255 Points: 115 # Comments: 21

now uses updated weights on , with lower pricing for input, cached input, and output tokens.Mistral Large 4 from MistralAI Gateway According to , the updated weights improve agentic coding and defensive cybersecurity, with greater token efficiency on coding tasks. The release also improves spreadsheet creation and editing, web research, and resistance to prompt injection.Mistral’s release announcement If you were already using via AI Gateway, there is no need to change your model to start using