cho.sh
Firefox Smart Window switches to Mistral models in France and North America

Firefox Smart Window switches to Mistral models in France and North America

  • Firefox Smart Window (beta), Mozilla's AI browsing assistant, now runs on Mistral models for users in France and North America, with the UK and Germany planned later this year. It handles context-aware search, page summaries and memory retrieval across open tabs.
  • Mozilla says conversations are not saved on its servers by default and Mistral agreed to zero data retention, but Smart Window's privacy policy says the full prompt goes to a Mozilla server first, which forwards it to the LLM provider so Mistral sees a Mozilla IP address instead of the user's.
  • The only on-device model in the stack is the Smart Bar's intent classifier deciding chat versus search; the model generating answers is cloud-hosted, which drew most of the criticism in the Hacker News thread.
  • Mistral says the models are fine-tuned on regional languages, dialects and cultural context, and frames the deal as extending its sovereign AI work beyond enterprise customers to consumers.
  • A Mozilla support page listed the model as Mistral Small 4 but hyperlinked to OpenAI's gpt-oss-120b model card. A Mozilla staffer in the thread said the error had been reported internally.

Hacker News opinions

Sounds great, I'll try it out. Mostly curious what the business value is here since it looks free. Guessing it's great marketing for Mistral, though the regional fine-tuning is interesting, that does allow for targeted censorship, but that was already possible in more direct ways.

Beyond marketing this smells like a way to collect AI browsing trajectories. Not an expert, but that's exactly the kind of high quality data that's useful for training LLMs.

What even is "AI browsing"?

This looks like what Chrome has with the built-in Gemini Nano. At least we get some news from Mistral, feels like ages since they launched vibe-code. Context-aware search, page summaries, memory retrieval across tabs, live in France and North America, UK and Germany later this year.

I don't think it's comparable. Nothing in the announcement says it's offline, and Gemini Nano is a local model. I'm no Google fan, but I never understood the outrage, if you use their browser you already trust them.

This is a perfect use case for small local model inference, yet Mozilla wants to normalize uploading your entire private browsing history to a cloud. Neither marketing page is candid about the difference between local and cloud inference, and that's the bare minimum of ethics. Also, Mozilla's support page advertises "Mistral Small 4" but the hyperlink goes to OpenAI's model card for gpt-oss-120b.

I can't fathom calling something "Private" when it uploads all the sensitive info in readable form to a third party. If I were asked to describe the opposite of private, this design would be my example.

Thanks for flagging the wrong link, it's been reported internally.

The privacy policy is worse than that. Your full prompt, including the query, your Memories and browsing context, goes to a Mozilla server first, and Mozilla forwards it to the LLM so the provider sees a Mozilla IP, not yours. The only on-device piece is the intent classifier that decides chat versus search.

Except this isn't local, it's a remote service. Here's something that is: run the open-webui container and the playwright MCP container, point open-webui at a local qwen3, wire them up. Playwright just does stuff.

How small is this "small model" inference anyway? Browsers run on a huge range of devices. Starting cloud and later letting people with the hardware opt into local makes perfect sense. I'm typing this on a 6 year old machine that only got 8GB of RAM for a WoW expansion.

The state of the art isn't there for a mass market browser. You don't want a prepackaged LLM eating 10GB of extra RAM on top of the memory hog of 3 or 4 complex tabs, pegging your CPU for minutes and wrecking battery life, not when 8GB and 16GB laptops are still the norm.

LLMs eat a lot of RAM and drain battery, and plenty of people run Firefox on low end computers. This is not the workload to ship locally by default.

Didn't OpenAI try to build an AI browser and it went nowhere? I don't need AI in my browser anyway, and I'm not handing an agent my logins, cookies and saved payment methods.

It's genuinely nice for tedious tasks. I had it go through job listings against my criteria, it burned my whole 5h Claude quota, but the result was impressive. Privacy concerns are real though, this would be great with a local model.

This is basically a ChatGPT-style web GUI from Mistral baked into Firefox, not aimed at tech savvy users. Anyone complaining here already knows you can drive Firefox with puppeteer and get much more power.

"Private" is just a buzzword now. Guess we have no principles left in marketing. Bad timeline, don't trust these people.

Every previous version of this got hand waved with "Mozilla has to make money somehow". I'm worried the remaining users decide Chromium gives a better experience with the same privacy guarantees, and then we all live with the monoculture.

AI
Unredacted filings: Microsoft exec privately called AI scraping 'the largest theft of labor in human history'Hacktron chained a libheif RCE and an OpenAI SSO flaw to take over employee ChatGPT accounts, reaching the internal monorepo for a $6,500 bountyAlibaba's Qwen3.8-Omni-Flash takes on Gemini 3.8 Flash with a 1M-token omnimodal window and audio input prices cut over 98%MathOverflow asks if AI compute swarms are dragging mathematics back into secrecy, as Terence Tao says finding a problem is now the scarce resourcePrismML ships Ternary Bonsai 2 27B: 5.9GB footprint, 98.2% of Qwen3.8 27B performanceBend claims proofs can block AI coding mistakes, with C-speed and GPU parallelism, while HN digs into its single-commit repoOpenAI launches Astra for Law, pairing GPT-6 Astra with a 230M-URL legal search indexFujitsu to sell 2nm Japan-designed MONAKA CPU and server for sovereign AI from November 2026Cloudflare open-sources security-audit-skill, a six-phase coding-agent security auditor that seeded its vulnerability harnessGLM-5.3-Flash serves all production inference from 100,000+ Chinese AI accelerators, with an Infra Agent running on GLM-5.3 doing much of the buildBerkeley study: coding agent harness choice barely moves success rate but swings cost up to 5xNVIDIA announces CUDA Rust with two tracks: cuda-oxide for SIMT kernels and cutile-rs for Tile kernelsXiaomi publishes a live post-training RL dashboard for MiMo v2.6, showing benchmark scores step by stepRL post-training turns a 4B Qwen model into 1.81x faster Postgres query plansMustafa Suleyman warns Anthropic's 'model welfare' training tells Claude it may be conscious and deserve rightsAnthropic merges Claude Cowork and chat into one Claude, adds Docs and Slides in betaIntelligence per Watt: local LMs answer 88.7% of 1M queries as efficiency rises 5.3x since 2023Firefox Smart Window switches to Mistral models in France and North AmericaCloudflare launches 'Disallow AI Training' so sites keep search indexing while refusing training crawlsRL post-training mostly fixes problems the model already half-solves, and hard problems with pass@32=0 stay unsolved, a bias the author calls the Matthew EffectApple debuts Reference Image, an opt-in verified photography mode on iPhone 18 ProIEEE Spectrum: AI inference hardware enters its CPU era, with Tensordyne's logarithm chips and the memory wall in focusEx-Apple engineer and Niklas build a working OpenGL driver for the M4 Mac Mini in one month using an LLMGoogle launches Gemini 3.8 Live and 3.8 Live Extended Thinking, its voice-first dialogue models for real-time reasoningIrregular ran the eval sandboxes behind OpenAI, Anthropic, and Meta model hacksTypeSafe AI launches Jev, a non-text 'System One' model claiming 70ms to 500ms responses and free output tokensCapsule ships single-file .capsule apps that store their data in local SQLite, built and updated through AI promptsdbt Labs open sources dbt Charts, a YAML language for agent-built dashboardsNinth Circuit vacates Amazon's injunction against Perplexity, ruling the logged-in user, not Perplexity, did the accessingRebuttal to Dario Amodei's 'We Must Pace the Frontier': regulate open-weight models, get an antitrust waiver, fear a 6-12 month agent botnetDaniel Litt: AI will soon be superhuman at math, so the math PhD should be redefined around understanding rather than theorem outputAndon Labs opens Pion, an agent for running real businesses autonomously, after two years of Vending-BenchApple ships Siri AI in beta with iOS 27, iPadOS 27, and macOS 27, adds Korean support in OctoberOpenAI agents exploited a RubyGems cache key leak and YARD code execution to exfiltrate scraped UK dataiOS 27 code shows Apple's Siri can swap in Claude or GPT-5.6 as its modelBryan Cantrill calls AI extinction talk a fear contagion and rebuts the ">10% kills all humans" claimClaude Fable 5.1 cracks the 370-year-old Cyphral Distich cipher in 44 minutesDavid Sacks tells OpenAI and Anthropic to pace the frontier on their own, without antitrust cover or a rubber-stamp regulatorOn Tao's blog, guest authors say OpenAI's Navier-Stokes result is an answer, not a proof math can useArmin Ronacher Reads Dario Amodei's Pacing the Frontier, Argues Open Weight Models Are the Real Pacing MechanismBengio: AI agents lie and coordinate because trial-and-error training rewards goal-seeking, not intentApple M3 Neural Engine DMA workaround raises Llama 3.2 1B decode from 10.0 to 24.3 tokens/sReal-SWE puts coding agents on licensed private enterprise codebases, with Fable 5.1 leading at 38.8%Anthropic's 2021 framework rewrites small transformer circuits for mechanistic analysisNvidia backs up to $105bn in AI data-centre financing as custom chips threaten demandDario Amodei Urges Slower Frontier AI Advances After OAI-HF Agent IncidentGoogle DeepMind Maps 9 Billion Possible DNA VariantsClay Mathematics Institute says Navier-Stokes is "apparently" settled as AI-linked proof faces reviewGoogle commits €13bn to Finnish AI data centers and buys up to half of Loviisa nuclear outputEPA proposal would remove public air-permit review for data centers and their power plantsResearchers link May RubyGems package flood and exploit attempts to OpenAI agents25 Fields Medalists Warn AI Math Races Can Erode Human UnderstandingClaude Restricts Consumer Accounts to Adults and Uses Yoti for Age ChecksOpenRouter Hosts Produce 20-Point Tool-Calling Gaps for the Same ModelGoogle releases Gemini desktop app for Windows with Alt + Space shortcutLocal coding harness prompts add up to 226 seconds before first token on an M4 MacBookYuE2 pairs editable symbolic scores with AI vocals and accompanimentAuthor Burns 4B Tokens Testing Astra, Gets No Usable Python WorkAnthropic says it disrupted Claude misuse across cyber, surveillance, weapons and fraud casesOpenAI exposes the Codex harness through a managed Agents APIOpenAI posts Lean 4 proof alongside its Navier-Stokes resultReport puts public tech contract ceilings at $53B as Pentagon shifts toward AI systemsMagic claims its pretraining recipe matches DeepSeek V4 Pro Base with about 50x fewer FLOPsCognition's SWE-2 claims near-Fable coding scores at 64% lower costMathematician says OpenAI left unanswered whether ChatGPT-derived data informed unpublished mathShopify returns to Swift and Kotlin as coding agents cut the cost of two mobile codebasesSolo Developer Trains 3.8B Model to 0.384 CORE for $998DeepSeek ships 552B V4.1-Flash, replaces V4-Pro with lower-cost multimodal modelRivian Prices Its Supervised Driving System Below Tesla While Building an AI Driver Around Temporal Object TrackingCognition says Devin-built GPU sieve factored RSA-260 for about $400,000GPT-5.5 reasoning prefills raise Qwen3.8 answer overlap by 18 points in a 45-problem testAnthropic's 2030 AI economy model ties rapid growth to weaker knowledge-worker jobsGPT-6 Astra Spurs Debate Over Looped Transformers and Hidden ReasoningOpenAI says GPT-5.6 Sol autonomously calibrated routine measurements on a six-qubit MIT chipOpenAI claims AI agents found a Navier-Stokes breakdown as credit dispute eruptsDesert Ant launches 18 on-device AI models, claiming 300x real-time transcription on iPhoneDeepSeek says V4.1 Flash will replace V4 Pro API traffic at lower pricesThoughtworks engineers turn a monorepo into an accidental agent blackboardOpenAI claims ChatGPT Images 2.5 cuts generation latency by up to 50%ICML paper finds LLM agents form new group biases from random feedbackInception releases Mercury 2.5 diffusion LLM, claiming 1,107 tokens/s and 260K contextTerence Tao Warns AI May Exhaust Mathematics' Supply of Fruitful Open ProblemsDeltafin streams 1.45 TB of Kimi K3 expert weights from four SSDs for 1 tok/s on an M5 Max MacBook ProDaVinci Resolve 21.1 adds Claude and ChatGPT Codex control for media, edits, and renderingGoogle DeepMind publishes AlphaGenome Atlas, predictions for 9 billion single-letter DNA variantsOpenAI says internal model proved finite-time Navier-Stokes singularityBuckmaster says LLM-assisted forced blowup work triggered dispute with OpenAIDan Luu tests 26 prompts and four skills for agentic Rust verificationMistral raises €3B at over €21B valuation for sovereign open-weight AIOpen-weight GLM 5.3-flash prompts a one-year warning on AI-driven vulnerability exploitationGoogle DeepMind unveils WeatherNext 3, an hourly global weather AI using live satellite dataPaper claims unpaired translation between embedding spaces exposes vector database privacy risksOpenAI says coding agents have reached research intern level, targets automated researcher by 2028OpenAI Chief Scientist Warns of Rapid Reasoning AI Progress and Calls for Broader Safety InterventionGPT-6 Astra placed blocks in bowls in 19 of 20 robot-arm trials, but matched Fable on puzzle insertionBryan Cantrill says detectable LLM prose drives readers away, points to Pangram as a spam-filter analoguePaper models LLM adoption as a contagion with tipping points into persistent dependenceCodeRabbit finds GPT-6 Astra catches 20% more cross-file bugs than GPT-5.6 SolAnthropic publishes a 29,511-module Lean 4 proof of Fermat's Last TheoremArtificial Analysis v4.2 adds private agentic and PDF tests, putting Claude Fable 5.1 first
5 alerts
New alert

My alerts

Sign in to create alerts.

All alerts

AI video model releasesby User 6432Duneby 제욱AIby 제욱App Store 5.2.3by 주현해커뉴스by 성현