A Guth Labs publication ↗

Prime Radiant and Simon Willison released three prototype sandbox servers for coding agents, separating the agent itself from the environment where its commands execute. Willison described the release in a November 17 post. The project announcement describes implementations using WebAssembly and Node.js, Apple Conta…

Sources: blog.primeradiant.com, simonwillison.net

Models

More models →

Agents

More agents →

Tools

More tools →

Research

More research →

Business

More business →

Infrastructure

More infrastructure →
All updates →
  1. Models

    MountAIn raises €7.3 million for computer vision on low-cost chips

    MountAIn announced €7.3 million in seed financing on November 17 to commercialize software that runs computer vision applications on inexpensive, low-power processors. The Grenoble company named 360 Capital, Supernova Invest and Crédit Agricole Al…

    Sources: globenewswire.com
  2. Models

    Model ML raises $250 million for financial-services AI agents

    Model ML announced a $250 million Series B on November 17 at a $1.5 billion valuation, financing further development of AI agents for financial and professional services. General Atlantic and QED Investors co-led the round, which brings total fund…

    Sources: generalatlantic.com, techcrunch.com
  3. Models

    AssemblyAI introduces STT-3.7 models for live conversations, agents and recordings

    AssemblyAI introduced its STT-3.7 transcription family, with separate models for live conversations, voice agents and recorded audio, in an announcement now dated November 18. All three support 32 languages with automatic detection. The Realtime m…

    Sources: assemblyai.com
  4. Models

    AWS adds managed Dynamo and cache-aware routing to SageMaker inference

    AWS announced managed NVIDIA Dynamo support for SageMaker HyperPod Inference and a separate cache-aware routing option for SageMaker AI real-time endpoints on November 17. The Dynamo integration supports models spread across multiple machines and …

    Sources: aws.amazon.com
  5. Agents

    Vercel makes eve 1.0 agent framework generally available

    Vercel announced the general availability of eve 1.0 on November 17, moving its open-source framework for building AI agents out of the public preview that began in June. The framework provides the software loop that calls models, runs tools and d…

    Sources: vercel.com
  6. Agents

    Voyage Labs raises $12 million for Atlas travel agent

    Voyage Labs announced $12 million in seed funding on November 17 to develop Atlas, a specialized AI assistant for travel and local outings. Kleiner Perkins led the round, with SignalFire and SV Angel participating, according to the company release…

    Sources: prnewswire.com, techcrunch.com
  7. Models

    Suno adds image and video creation with Visuals

    Suno launched Visuals on November 17, adding image and video creation to its music platform on the web, iOS and Android. Users can start with a text prompt or reference image, create artwork, animate scenes and make characters sing or speak along …

    Sources: suno.com, help.suno.com
  8. Agents

    LoanPro details its Bedrock-powered loan servicing assistant

    LoanPro and AWS published a November 17 case study describing a production AI assistant that summarizes borrower accounts and recommends next steps for loan servicing staff. Built with AllCloud and the AWS Generative AI Innovation Center, the syst…

    Sources: aws.amazon.com
  9. Agents

    Meta brings its Muse personal agent to Australia

    Meta made Muse available in Australia on November 17, expanding access to its personal AI agent while local service integrations roll out progressively. The company announcement lists availability through iOS, Android, the web and a Mac app; suppo…

    Sources: about.fb.com
  10. Models

    LangChain adds granular access controls for sensitive LangSmith traces

    LangChain added three access controls to LangSmith on November 17, allowing enterprises to restrict sensitive trace content while retaining information engineers use to debug agents. The announcement says the controls are available on Enterprise p…

    Sources: langchain.com
  11. Agents

    GitHub Copilot app adds Azure DevOps preview and Enterprise Server support

    GitHub expanded its Copilot app on November 17 with an Azure DevOps integration in public preview and generally available GitHub Enterprise Server support. The release notes distinguish both from cloud sessions, which remain unavailable for those …

    Sources: github.blog
  12. Models

    GitHub expands Copilot Fusion to web, Visual Studio and mobile betas

    GitHub expanded Copilot Fusion on November 17 to github.com, Visual Studio and the GitHub Mobile beta apps for iOS and Android. The update says production mobile-app support is coming later, so the mobile rollout is not yet universal. Fusion selec…

    Sources: github.blog
  13. Agents

    GitHub previews live pull-request events without a webhook server

    GitHub introduced an Event Streams preview on November 17, allowing agents and other automation to receive pull-request updates over an open connection instead of repeatedly polling the service. Access is available on all GitHub plans after approv…

    Sources: github.blog
  14. Guth News

    GitHub previews adaptive code reviews and secret detection

    GitHub added an Adaptive effort option to Copilot code review on November 17, letting the reviewer choose Lite, Balanced or Max based on a pull request's complexity and risk. The public preview shows which level was selected and respects the maxim…

    Sources: github.blog
  15. Agents

    GitHub previews repeatable tests for Copilot skills, agents and tools

    GitHub opened Copilot Evals in public preview on November 17, giving developers a way to test how changes to skills, custom agents and MCP tools affect their Copilot setup. The release supports Copilot Individual, Business and Enterprise plans. Us…

    Sources: github.blog
  16. Models

    OVHcloud announces Olala mini, a compact model aimed at local AI

    OVHcloud announced Olala mini 7A1B on November 17, introducing a compact language model with seven billion total parameters and one billion active parameters. The company schedules its formal launch for the OVHcloud AI Summit on November 19. The a…

    Sources: corporate.ovhcloud.com
  17. Agents

    Perplexity adds a Multiagent setting with up to 16 collaborating agents

    Perplexity added a Multiagent effort level to Computer on November 17, letting a lead agent coordinate up to 16 supporting agents that share findings and check each other's work. The feature is available to Max subscribers on the web and sits abov…

    Sources: perplexity.ai
  18. Models

    Perplexity launches Orchestrator API with shared billing and model failover

    Perplexity launched its Orchestrator API on November 17, combining access to proprietary and open-weight models with consolidated billing and provider failover. Its announcement lists support for OpenAI, Anthropic, Google and xAI models alongside …

    Sources: perplexity.ai
  19. Models

    GitHub details the protocol behind cross-device Copilot sessions

    GitHub published an engineering account of its Agent Host Protocol on November 17, explaining how Copilot sessions continue across editors, desktop apps, phones and cloud environments. The post describes an open standard developed first in VS Code…

    Sources: github.blog
  20. Models

    GitHub says AI security pilot produced 99 checked reports and 55 fixes

    GitHub and Microsoft reported on November 17 that an AI-assisted security pilot produced 99 privately submitted vulnerability reports across six open-source projects, with maintainers fixing 55 by November 15. Fourteen fixes had reached a release,…

    Sources: github.blog
  21. Agents

    Dreamteam raises $53 million for a CRM built around four AI coworkers

    Dreamteam announced $53 million in funding on November 16 to expand customer relationship management software built around shared customer context and specialized AI agents. A company post dated November 17 says the total spans its seed and Series…

    Sources: dreamteam.ai, globenewswire.com
  22. Robotics

    Genesis AI raises $500 million to scale its general-purpose robots

    Genesis AI announced a $500 million Series A on November 17 at a $3.5 billion post-money valuation, with plans to expand training and deploy its Eno robot at scale. Premji Invest, Khosla Ventures and Eclipse co-led the round. Participants included…

    Sources: prnewswire.com, techcrunch.com
  23. Models

    OpenAI and Chile agree to explore public-sector AI training and startup support

    OpenAI and Chile's government announced a partnership on November 17 to explore AI use in public administration, civil-service training and support for local startups. The OpenAI for Chile announcement says the Ministry of Science, Technology, Kno…

    Sources: openai.com
  24. Models

    HeyGen launches Avatar 6 with script-aware gestures and two-person scenes

    HeyGen launched Avatar 6 on November 17, adding gestures that respond to a script and support for two avatars speaking in the same scene. Its product page describes controls for emotion, expression and movement, alongside API and MCP access. The c…

    Sources: heygen.com
  25. Agents

    Mirelo launches an agent for editing video sound with prompts

    Mirelo launched Sound Agent on November 17, adding prompt-driven audio editing to Mirelo Editor while leaving a video's pictures unchanged. The company's announcement describes a workflow in which users upload a video, request changes and review a…

    Sources: mirelo.ai
  26. Agents

    AWS expands its Forward Deployed Engineering service to 15 countries

    AWS has expanded its Forward Deployed Engineering service to 15 countries, making its embedded AI engineering teams available more broadly to commercial and public-sector customers. In a November 17 announcement, AWS said the service now covers th…

    Sources: aws.amazon.com
  27. Models

    Proton launches Lumo desktop apps with connectors, schedules and skills

    Proton released Lumo Desktop for Windows, macOS and Linux on November 17, adding app connections, scheduled tasks and reusable instructions to its AI assistant. The company’s announcement introduces three features: Connectors, Schedules and Skills…

    Sources: proton.me
  28. Agents

    Zenity expands Microsoft agent monitoring and Sentinel security alerts

    Zenity expanded its Microsoft agent-security coverage on November 17, adding monitoring across coding, office and employee-built assistants and upgrading its Microsoft Sentinel integration. The company announcement describes different levels of pr…

    Sources: streetinsider.com, businesswire.com
  29. Agents

    Knotch launches Ace for Agents to make website content readable by AI assistants

    Knotch launched Ace for Agents on November 17, adding a service that prepares companies’ existing website content for AI search and personal assistants without replacing the pages human visitors see. The product sits within its Agentic Content Eng…

    Sources: prnewswire.com
  30. Guth News

    Cloudflare adds optional file summaries to AI Search indexing

    Cloudflare added optional AI-generated file summaries to AI Search indexing on November 17, letting the service create descriptions, topics and keywords for text documents that produce at least three chunks. The release note says the feature is of…

    Sources: developers.cloudflare.com
  31. Models

    Cerebras puts combined GPU and wafer-scale inference into production

    Cerebras has put a split inference system into production that uses AMD Instinct GPUs to process prompts and Cerebras CS-3 systems to generate responses, the company announced on November 17. The arrangement separates two stages of running an AI m…

    Sources: cerebras.ai
  32. Agents

    Microsoft previews managed agent browsing in Edge for Business

    Microsoft introduced a managed agent browsing preview in Edge for Business on November 17, giving organizations a way to run Microsoft and third-party agents inside a browser governed by existing company policies. According to Microsoft’s Ignite a…

    Sources: blogs.windows.com
  33. Agents

    Microsoft starts Agent 365 Business Analytics rollout in Frontier preview

    Microsoft began rolling out Agent 365 Business Analytics in Frontier preview on November 17, adding tools for comparing AI spending, usage and business results across Microsoft and connected third-party services. The Ignite announcement says the s…

    Sources: techcommunity.microsoft.com
  34. Models

    OpenAI launches GPT-6.2 Astra and Sol with lower Astra API prices

    OpenAI released GPT-6.2 Astra and GPT-6.2 Sol on November 17, bringing its latest models to ChatGPT Work, Codex and the API. Astra targets demanding professional and coding projects, while Sol is positioned for everyday work. Both are available to…

    Sources: openai.com, deploymentsafety.openai.com +1
  35. Models

    ioMoVo V2 adds an agent builder and multi-model routing

    ioMoVo released V2 of its enterprise AI platform on November 17, adding a workspace for building agents and controls for choosing which AI models handle their requests, according to its announcement. The new ioStudio workspace lets teams create an…

    Sources: prweb.com
  36. Models

    Microsoft releases MAI Code 2 and expands its specialist AI lineup

    Microsoft announced new coding, security, image and speech models at Ignite on November 17, led by MAI Code 2 and its smaller Flash variant. The company describes the broader launch as a family of seven new models, emphasizing lower operating cost…

    Sources: microsoft.ai
  37. Agents

    Microsoft releases Copilot Home and Office, starts Autopilot rollout

    Microsoft made Copilot Home and Office in Copilot generally available on November 17 and began rolling out its Autopilot agent to Frontier customers, according to its Ignite announcement. Home combines Chat and Cowork in one starting point. Office…

    Sources: microsoft.com
  38. Agents

    Pedestal AI raises $4.7 million for supply-chain agents

    Pedestal AI emerged from stealth on November 17 with $4.7 million in seed funding to build AI agents for supply-chain and logistics operations. Eclipse led the round, joined by RRE, Converge VC, Timescale VC and VitalStage Ventures, the company sa…

    Sources: prnewswire.com
  39. Agents

    Scotiabank launches two AI agents for commercial banking teams

    Scotiabank launched Commercial Navigator on November 17 for its entire Canadian Commercial Banking team, adding two AI agents to support credit-related work. The tool is integrated into existing systems and powered by Scotia Intelligence, accordin…

    Sources: newswire.ca
  40. Models

    Argentum and Kyndryl sign $630 million agreement for B300 GPU capacity

    Argentum AI and Kyndryl announced a 20-year, $630 million data-center collaboration on November 17, covering an initial 10 megawatts at Kyndryl’s Boulder, Colorado facility. Argentum is deploying NVIDIA B300 systems there for AI workloads, accordi…

    Sources: prnewswire.com
  41. Agents

    Microsoft previews expanded Entra controls for local AI agents

    Microsoft announced expanded Entra controls for AI agents in a November 17 Ignite post, including previews for discovering local agents and adjusting their access as risk changes. The announcement says discovery covers 60 local agent types, includ…

    Sources: techcommunity.microsoft.com, learn.microsoft.com
  42. Tools

    Elementl uses Google Gemini to plan nuclear projects

    Elementl Power announced a collaboration with Google Cloud on November 17 to build AI tools for developing nuclear power projects. Its Lattice project-delivery suite uses Gemini for business and BigQuery, according to the companies’ announcement. …

    Sources: googlecloudpresscorner.com, prnewswire.com
  43. Agents

    Evoke launches lab for investigating AI agent incidents

    Evoke Security launched Defender CTF on November 17, a hands-on training lab for investigating incidents involving AI agents. The company announcement describes a staged incident modeled on behavior it has observed in customer environments. Partic…

    Sources: prnewswire.com
  44. Agents

    Ordway adds lockbox check matching to its cash application agent

    Ordway expanded its cash application agent on November 17 to match bulk bank lockbox deposits with invoices. The company’s announcement says the update analyzes individual check images and bank lockbox reports to reconcile payments that may appear…

    Sources: prnewswire.com
  45. Tools

    The Rx Assistant unveils no-code Builder for pharmaceutical AI services

    The Rx Assistant unveiled Builder on November 17, a no-code platform for pharmaceutical companies to develop and deploy AI interactions across voice, text messages, email and web chat. The company announced the product at HLTH 2026 as the infrastr…

    Sources: prnewswire.com
  46. Guth News

    Hyperbolic hedges $5 million of GPU rental revenue with Luxor and Architect

    Architect and Luxor announced on November 17 that they completed a privately negotiated option transaction to hedge Hyperbolic's NVIDIA B200 GPU rental revenue. The companies' release puts the covered future revenue at approximately $5 million in …

    Sources: prnewswire.com
  47. Guth News

    b.well launches health skills for record-based AI assistance

    b.well Connected Health launched health.skills on November 17, a library of written clinical workflows for AI assistants that consult a person’s health record. The library is available in its Health AI SDK and bailey assistant, according to the co…

    Sources: prnewswire.com
  48. Guth News

    Spara launches free Scout assistant for proactive sales outreach

    Spara launched Scout on November 17, a free AI assistant designed to identify sales opportunities and prepare outreach without waiting for individual prompts. The company says Scout uses customer conversations, email and CRM information to maintai…

    Sources: prnewswire.com
  49. Models

    Kaseya adds MCP servers and 17 skills for IT management

    Kaseya announced first-party Model Context Protocol servers and 17 prebuilt skills on November 17 to connect AI assistants with Autotask, Datto RMM and IT Glue. The servers are available through GitHub in early access, according to the company’s r…

    Sources: prnewswire.com
  50. Agents

    CoreStory releases V3 to give coding agents legacy business rules

    CoreStory released version three of its code-intelligence platform on November 17, adding a persistent map of legacy software’s business rules for AI coding agents. The company announcement says V3 connects those rules to the code that implements …

    Sources: businesswire.com, recorderonline.marketminute.com
  51. Agents

    Cogent releases DefenseOS for AI-driven vulnerability management

    Cogent released DefenseOS on November 17, expanding its cybersecurity platform with tools that give AI agents organizational context and automate recurring vulnerability-management work. DefenseOS and Agent Automations are available to all Cogent …

    Sources: prnewswire.com
  52. Guth News

    Argenx expands use of Cradle’s AI antibody discovery platform

    Argenx is expanding use of Cradle’s AI protein-engineering platform to more scientists and antibody discovery programs, Cradle announced on November 17. The expansion follows what the companies describe as successful initial work, although the ann…

    Sources: prnewswire.com
  53. Models

    Captello releases MCP connector for event data and updating dashboards

    Captello made its Model Context Protocol connector generally available on November 17, bringing event data into AI assistants including Claude, ChatGPT and Gemini. The company says customers can use conversational requests across lead capture, mee…

    Sources: prnewswire.com
  54. Models

    Isomorphic Labs raises $5 billion to advance AI-designed medicines

    Isomorphic Labs announced a $5 billion Series C funding round on November 17 to expand its AI drug-design technology and advance potential medicines toward clinical trials. Greenoaks and Sequoia Capital co-led the financing, according to the compa…

    Sources: prnewswire.com, isomorphiclabs.com +1
  55. Agents

    Daydream launches shopping agent with virtual try-on and approved checkout

    Daydream launched an AI fashion agent on November 17 that can assemble outfits, provide virtual try-ons and complete purchases with a shopper’s approval. The service is available through its iOS app and text messaging, according to the company’s a…

    Sources: prnewswire.com
  56. Guth News

    OpenAI commits £1 million to King's Trust AI-skills partnership

    The King's Trust and OpenAI announced a £1 million partnership on November 16 to help young people in the United Kingdom develop practical AI skills for work. The program targets up to 25,000 participants over 18 months, prioritizing people outsid…

    Sources: openai.com
  57. Agents

    Amazon Connect adds interactive controls to AI agent assistance

    AWS announced on November 16 that Amazon Connect Customer agent assist can respond with cards, carousels, buttons and input fields, rather than relying only on text. The controls let contact-center employees review information and supply or confir…

    Sources: aws.amazon.com, docs.aws.amazon.com
  58. Agents

    Google adds AdMob agent skills and previews a monetization assistant

    Google announced a set of AdMob AI tools on November 16, separating coding help available now from account-level assistants still being tested. Google announcement The available agent skills are reusable instructions for tasks such as migrating to…

    Sources: blog.google
  59. Models

    Meta says AI screening helped reduce reports of Instagram recommendations in India

    Meta published a November 16 update on changes to Instagram recommendations in India, saying users now report recommended content about half as often as a year earlier. The company attributes the improvement to stronger screening and systems tuned…

    Sources: about.fb.com
  60. Agents

    AWS expands long-running AgentCore runtime instances to six more regions

    AWS announced on November 16 that AgentCore runtime instances are available in Northern California, Canada Central, London, Paris, Stockholm and Seoul. The expansion concerns dedicated runtime instances, not the initial launch of AgentCore. AWS an…

    Sources: aws.amazon.com
  61. Models

    SageMaker adds live cache-aware routing for multi-instance LLM endpoints

    AWS announced on November 16 that SageMaker AI real-time endpoints can route requests according to the prompt content already cached on each serving instance. The aim is to avoid recomputing long, repeated inputs when traffic spans several GPUs. A…

    Sources: aws.amazon.com
  62. Agents

    S3 Vectors combines semantic and keyword retrieval in one query

    AWS announced on November 16 that Amazon S3 Vectors now combines vector similarity and full-text search in a single query. The release targets retrieval-augmented generation, agent memory and search applications. AWS announcement Developers attach…

    Sources: aws.amazon.com
  63. Agents

    AWS adds identity delegation and session replay to WorkSpaces agents

    AWS announced new WorkSpaces capabilities on November 16 for setting up and supervising AI agents. This extends a service launched earlier in the year, rather than introducing WorkSpaces agents for the first time. AWS announcement Coding assistant…

    Sources: aws.amazon.com
  64. Models

    Gemini separates chat history from model-training permission, with a retention caveat

    Google announced on November 16 that new Gemini settings will let users retain their chat history while opting out of generative-AI model training. Users will still be able to choose history-retention periods and manually review or delete saved co…

    Sources: blog.google
  65. Guth News

    Gemini Notebook adds interactive sites built from users’ sources

    Google announced on November 16 that Gemini Notebook can transform a user's sources into interactive sites. The feature begins rolling out globally on the web for users aged 18 and older, including Google Workspace subscribers; mobile support is d…

    Sources: blog.google
  66. Tools

    Sabre plans Grok travel preview with booking and disruption-response tools

    Sabre announced on November 16 that it would preview a Grok Bot travel integration at its November 17 investor day. The planned Mosaic Travel Connector is intended to let people search flights and hotels, book trips and manage reservations through…

    Sources: prnewswire.com
  67. Models

    GitHub retires three Copilot models, with limited exceptions

    GitHub confirmed on November 16 that Claude Opus 4.8, Claude Opus 5 and GPT-5.6 Terra have been deprecated across most Copilot experiences. The change implements a schedule announced on October 15, rather than a new surprise deadline. GitHub chang…

    Sources: github.blog
  68. Models

    du bundles sovereign AI inference and agent services in the UAE

    UAE telecom operator du announced National HyperAI on November 16, combining AI inference, GPU rental, hosted models and agent workflows under one commercial agreement. The company says the service is available immediately to UAE government and en…

    Sources: zawya.com
  69. Guth News

    Google rolls out 66 regional voices across the Gemini mobile app

    Google announced on November 16 that it is rolling out 66 new regional voice options in Gemini on Android and iOS. The set includes U.S., U.K. and Indian English, alongside Hindi, Brazilian Portuguese, Indonesian and Korean options. This is a mobi…

    Sources: blog.google
  70. Agents

    Databricks opens an AI Index that compares coding-agent cost with performance

    Databricks launched a public AI Index on November 16, combining anonymized Unity Gateway usage with coding-agent evaluations on its own engineering tasks. The company says the resource draws on more than 10,000 companies and will receive weekly up…

    Sources: databricks.com, developers.databricks.com
  71. Models

    Baseten previews a voice-agent stack that brings models and orchestration together

    Baseten announced its Voice Agent Stack and Realtime Voice on November 16. Both are in private preview, not general availability. The stack hosts an agent's orchestration alongside its speech and language models, while letting components scale sep…

    Sources: baseten.co
  72. Models

    Noeri launches with $110 million to build enterprise AI that keeps learning

    Noeri, founded by Stanford professors Yejin Choi and Carlos Guestrin, launched publicly on November 16 with $110 million in funding. Menlo Ventures led the round, with investors including Madrona, NVIDIA, Cisco Investments and Snowflake. The finan…

    Sources: businesswire.com, noeri.ai +1