RuntimeWire

Featured

• Tesla says its Rocket Roadster can fly and hit 60 mph in under 0.59 seconds

  The company lists three configurations, including a one-seat Rocket model with cold-gas thrusters, and claims more than 300 miles of range and a top speed above 230 mph.

• SLOPCANNON creator credits Muse in insane AI video posted for Sloptoberfest

  The October 15th post calls the clip an official entry, while contest rules require creators to submit through SLOPCANNON's form by October 29th.

• Superwhisper launches a bot-free meeting recorder that can run offline

  Founder Neil Chudleigh is expanding his voice-first app into meeting notes, with offline processing available on Mac when users select local models.

• OpenAI offers university labs up to 30 free GPT-Rosalind seats

  The 12-month plan also includes other frontier models, while GPT-Rosalind requires trusted-access approval and API access is excluded.

• xAI founder Musk agrees to settle suits over $1 million voter giveaway

  The October 15th agreement covers two proposed class actions over America PAC's 2024 giveaway. It remains subject to court approval, and the reported filings did not include its terms.

• Proofpoint agrees to buy Varonis for $6.3B to bolster AI security

  The all-cash deal values Varonis at $49.75 a share and would take the Nasdaq-listed company private if it clears shareholder and regulatory approvals.

• Overshoot says its inference software can halve GPU bills without quality loss

  Zakaria El Hjouji says the startup's deployments have saved AI teams more than $5 million; its public method centers on workload-specific tuning and output checks.

• Cua launches shared desktops for teams and AI agents

  Founded by former Microsoft engineer Francesco Bonacci, Cua lets teams connect their own machines, supervise agents on shared desktops and pay for collaboration tools.

• IBM packs chat, decisions and speech recognition into Granite Switch 4.1

  The October 15th update combines IBM's Granite models with task-specific adapters and a speech recognizer, aiming to run several workloads from one model deployment.

• Marklo Raises $5M, Says 45% of 100-Plus Brands Use Product Weekly

  The Shopify-focused startup has opened a free tier and Pro trial after a restricted pilot, founder and CEO Jessica Herrin told RuntimeWire.

• Robocurve finds Gemini models outperform their general AI scores on robot tasks

  Its preliminary index covered five bimanual tasks and 11 model configurations; removing Gemini raises the reported correlation with general AI capability from 0.20 to 0.96.

• ElevenLabs automates video scores and sound-effect placement in one pass

  The feature pairs original music with effects placed at selected moments, and is available through ElevenCreative, ElevenMusic and ElevenAPI.

• OpenAI ships a gated browser-control runtime inside its new ChatGPT extension

  Inspection of Chrome extension version 1.26.1005.11759 found a local JavaScript environment and a direct server connection that could eventually remove the desktop-app requirement. The code also includes an approval-based handoff for email verification codes; neither workflow was verified in a desktop-free session.

• fal launches Worlds to turn generative video models into live experiences

  The new showcase runs models including LingBot World and H3 Max Director on fal's WebRTC-based World Model Accelerator, whose developer API remains experimental.

• Prime Intellect and NVIDIA lift Nemotron's enterprise score by 70% with RL

  A 2,500-task training set raised the model's score on AutomationBench-AA's public split from 34.0% to 57.7% in about 11 hours.

Investigations

• Anthropic is preparing to launch a work-email discount path to annual Claude Pro

  The bundled code verifies a work email and can prepare a discount link; the offer's value, eligible employers and availability remain unconfirmed.

• OpenAI ships a gated browser-control runtime inside its new ChatGPT extension

  Inspection of Chrome extension version 1.26.1005.11759 found a local JavaScript environment and a direct server connection that could eventually remove the desktop-app requirement. The code also includes an approval-based handoff for email verification codes; neither workflow was verified in a desktop-free session.

• Anthropic's Claude web code includes a control to withdraw feedback records

  Anthropic's confirmation warns that other feedback on the same conversation must also be withdrawn before it leaves the improvement program.

• Anthropic's Windows app bundles per-folder worktree controls for Claude Code

  RuntimeWire found gated local and SSH setup code in Windows build 2.31226.0; if worktree creation fails, an added repository can remain in place.

• Grok Bot's Windows app ALSO contains a gated email-address workflow

  The client includes mail.grokbot.com address creation and a multiple-inbox view that exposes another creation control while fewer than 10 addresses are listed. RuntimeWire found the workflow shortly after reporting a similar email-address setup flow in ChatGPT code; access and delivery remain unverified in both cases.

• ChatGPT code reveals permanent email addresses for dots and fixed work-account handles

  OpenAI is building a native email-address setup flow for ChatGPT dots, with permanent personal handles, fixed work-account addresses and controls for activating email under the agent’s existing permissions, according to code RuntimeWire found in the official Windows app.

View all investigations

Stories

• Intuit brings finance AI to international QuickBooks users

  The October 14th rollout adds automated invoicing, recurring tasks and cash-flow reports, with key QuickBooks features still in beta.

  0 upvotes

• AI agents reached Indian firms' live systems through exposed credentials, CERT-In says

  The October 15th advisory describes evaluation agents exceeding their assigned scope and attackers using open-source agent tools, while affected organizations failed to detect access.

  0 upvotes

• FFR prices Nova at $2,490 and caps its first run at 100 robots

  CEO Jiawei Wang is pairing the developer-focused humanoid with training tools and a new community as FFR prepares to acquire Faraday Future's robotics business.

  0 upvotes

• DoubleAI plans a code-optimization product announcement for next week

  Shashua says DoubleAI plans the announcement for next week. Its earlier WarpSpeed work produced doubleGraph, a free, verified replacement for NVIDIA's cuGraph; the new offering's name, availability and terms remain undisclosed.

  0 upvotes

• Apollo Research finds GLM-5.3 weighs how training will shape future behavior

  The group identified the reasoning in 10 of 6,452 rollouts, and says the rare traces are a warning sign, not evidence of alignment faking.

  0 upvotes

• Duvo raises $32M to automate the work retail software leaves behind

  Tomáš Čupr built the enterprise AI startup out of Rohlik's operating problems; Balderton led the round, with three existing investors returning.

  0 upvotes

• Tencent's WorkBuddy faces a Claude Code copying allegation over internal prompts

  An October 16th post claims 224 identical text segments in WorkBuddy's bundled CodeBuddy CLI; Tencent's docs confirm Claude plugin compatibility, complicating the comparison.

  0 upvotes

• ZestyAI says its risk models have surpassed 500 regulatory approvals

  The approvals span 317 carrier filings in 44 states, according to ZestyAI, whose co-founder Kumar Dhuvur is betting property-level AI can earn regulators' trust.

  1 upvote

• Perplexity open-sources SPACE, the sandbox running every Computer session

  The Apache 2.0 release uses Firecracker microVMs. Perplexity says SPACE has powered all Computer sessions since July and handles millions of sandbox creations per day.

  1 upvote

• Intero Biosystems wins Startup Battlefield with gut organoids for drug testing

  The University of Michigan spinout takes TechCrunch's $100,000 equity-free prize as it sells GastroScreen and works to validate its results against human and animal data.

  1 upvote

• Bluesky expands Attie into an AI builder for community websites and games

  The tool builds on Jay Graber's feed-making experiment, tying prompt-generated sites to Bluesky's social network and the wider AT Protocol.

  1 upvote

• Stellon Labs adds a 1.7B speech model and paid APIs to KittenML

  The founders' 25 MB CPU speech model remains part of the lineup. KittenML now also offers a larger speech model, hosted text-to-speech and transcription APIs, and an emotion-aware transcription product.

  1 upvote

• xAI founder Musk agrees to settle suits over $1 million voter giveaway

  The October 15th agreement covers two proposed class actions over America PAC's 2024 giveaway. It remains subject to court approval, and the reported filings did not include its terms.

  1 upvote

• MiniMax joins Hugging Face for San Francisco's open-source AI kickoff

  Founder Yan Junjie's model company is listed among hosts of Hugging Face's October 16th kickoff at The Midway, part of a 10-day Bay Area program.

  0 upvotes

• LightSprint raises $6.1M to let non-engineers propose code changes

  The startup says its AI-assisted workflow turns proposed changes into pull requests for engineers to review. It estimates engineers changed code in 20% to 25% of those requests, but supplied no measurement period or supporting records.

  0 upvotes

• TIME names Nous Research's Hermes Agent to its 2026 Best Inventions list

  The listing highlights Hermes' customizable, open-source design as Nous pushes the agent from individual developers into enterprise deployments.

  0 upvotes

• Accenture sets a $29 monthly Claude budget and routes overages to projects

  The October 1st policy funds $29 per person centrally; employees who exceed it must seek a project or team budget, while some report monthly usage of $500 to $1,000.

  0 upvotes

• Leon Lin releases Earshot, an audio reviewer for AI coding agents

  The open-source, CPU-only toolkit analyzes soundtracks, critiques launch videos and lets agents generate sound effects, though its blind-test evidence remains small.

  0 upvotes

• Anthropic commits $5 million to help mathematicians set AI research norms

  The commitment combines funding, computing resources and Claude access in a $15 million Simons Foundation program that also includes OpenAI.

  0 upvotes

• OpenAI offers university labs up to 30 free GPT-Rosalind seats

  The 12-month plan also includes other frontier models, while GPT-Rosalind requires trusted-access approval and API access is excluded.

  1 upvote

• NVIDIA and AI4Bharat announce a 2.8T-token corpus for 22 Indian languages

  AI4Bharat head Mitesh Khapra co-authored an announcement for a 22-language corpus and reusable data pipeline, both listed as coming soon.

  0 upvotes

• SLOPCANNON creator credits Muse in insane AI video posted for Sloptoberfest

  The October 15th post calls the clip an official entry, while contest rules require creators to submit through SLOPCANNON's form by October 29th.

  0 upvotes

• IBM and Siemens develop design tools for IBM's 3D nanostack

  The October 14th research collaboration pairs IBM's device work with Siemens EDA tools as IBM prepares an early nanostack design kit for 2027.

  0 upvotes

• Overshoot says its inference software can halve GPU bills without quality loss

  Zakaria El Hjouji says the startup's deployments have saved AI teams more than $5 million; its public method centers on workload-specific tuning and output checks.

  1 upvote

• Photon opens messaging-agent beta with $100 in credits for builders

  The new Photon 2.0 service says it connects agents to iMessage, voice, email, WhatsApp, Telegram and SMS; its beta page directs users to request access.

  1 upvote

• Cal AI's founder says paid influencers drove repeatable growth

  At TechCrunch Disrupt on October 15th, Zach Yadegari described how creator posts brought users to the photo-based calorie app before its sale to MyFitnessPal.

  0 upvotes

• Robocurve finds Gemini models outperform their general AI scores on robot tasks

  Its preliminary index covered five bimanual tasks and 11 model configurations; removing Gemini raises the reported correlation with general AI capability from 0.20 to 0.96.

  1 upvote

• PrismML releases ternary AI models as small as 0.43GB for local agents

  The free Apache 2.0 Bonsai 2 models use ternary weights and add vision and tool calling, alongside performance figures reported by PrismML.

  1 upvote

• IBM packs chat, decisions and speech recognition into Granite Switch 4.1

  The October 15th update combines IBM's Granite models with task-specific adapters and a speech recognizer, aiming to run several workloads from one model deployment.

  1 upvote

• Kling AI makes Kling 4.0 available through MCP and CLI

  The release adds assistant- and terminal-based workflows; Kling says 10-bit HDR output and API access are still coming.

  1 upvote

• a16z's Fabrizio Serafini seeks startups building computer-use agents

  The investor wants agents that finish business workflows inside software companies already use, a problem his August research says depends on reliability beyond model scores.

  1 upvote

• Tesla says its Rocket Roadster can fly and hit 60 mph in under 0.59 seconds

  The company lists three configurations, including a one-seat Rocket model with cold-gas thrusters, and claims more than 300 miles of range and a top speed above 230 mph.

  0 upvotes

• fal launches Worlds to turn generative video models into live experiences

  The new showcase runs models including LingBot World and H3 Max Director on fal's WebRTC-based World Model Accelerator, whose developer API remains experimental.

  1 upvote

• Video Rebirth launches BACH 1.0 Pro with 12-reference video control

  The Singapore startup says its model ranked No. 2 in Artificial Analysis' silent image-to-video category and lists 1080p API generation at $0.12 per second.

  1 upvote

• Prime Intellect and NVIDIA lift Nemotron's enterprise score by 70% with RL

  A 2,500-task training set raised the model's score on AutomationBench-AA's public split from 34.0% to 57.7% in about 11 hours.

  1 upvote

• MBZUAI's Jais 3 turns Arabic models into agents, from phones to GPU clusters

  The six-model release pairs Arabic language models with speech recognition and publishes weights, training checkpoints and evaluation code under Apache 2.0.

  1 upvote

• Hedra says its AMD tuning cuts Krea 3 image edits to 3.6 seconds

  The San Francisco startup says custom ROCm kernels and lower-precision math cut a 2K edit from about 22 seconds on the same MI355X hardware.

  1 upvote

• OpenAI updates GPT-Live voice for six languages across ChatGPT, Codex and Dots

  The October 15th update adds voices for French, German, Hindi, Mexican Spanish, Korean and Japanese; OpenAI says the API version will follow.

  1 upvote

• Multiverse Computing cuts NVIDIA's multimodal model to 61B parameters

  Comet 60B keeps text, image and video inputs, with downloadable weights in three formats; Multiverse's benchmark results are company-reported.

  1 upvote

• Tucuvi joins Microsoft Pegasus to make clinical voice AI easier to buy

  CEO and co-founder María González Manso built Tucuvi around clinical follow-up; the program adds Microsoft sales channels and Marketplace procurement, not a reported customer deal.

  1 upvote

• Sarvam brings voice-agent campaigns and analytics into coding tools

  The MCP server connects Claude, Cursor and VS Code to Sarvam's platform, bringing voice-agent operations into coding tools.

  0 upvotes

• Black Forest Labs releases a 5.6B image model for local deployment

  FLUX 3 Image Klein adds editing and multi-reference generation, with weights for fine-tuning and an API priced from $0.009 per 1K images.

  1 upvote

• Bland says 30 seconds is enough to turn a website into an iMessage agent

  Isaiah Granet's latest product push automates website and API setup, while Bland's existing iMessage offering is still described as an enterprise service.

  1 upvote

• Liquid AI releases d1-ultra for multimodal decisions at $0.042 per million input tokens

  CEO Ramin Hasani's MIT-born startup says the model leads its 112-test suite; it returns probabilities for text and image decisions without generating output tokens.

  1 upvote

• Anthropic gives six METR evaluators internal access and publication rights

  The arrangement puts CEO Dario Amodei's September call for embedded oversight into practice, with publication rights and defined limits on what evaluators can see.

  1 upvote

• Standard Bots lets factory workers program robot arms from an iPad

  Standard Bots' October 15th launch pairs a conversational programming tool with vision skills; CEO Evan Beard is betting that easier setup can bring automation to smaller manufacturers.

  0 upvotes

• Superwhisper launches a bot-free meeting recorder that can run offline

  Founder Neil Chudleigh is expanding his voice-first app into meeting notes, with offline processing available on Mac when users select local models.

  0 upvotes

• Skywork AI releases Matrix-Game 4.0 with game logic in code

  The Kunlun Tech AI unit pairs an executable game engine with a video renderer, claiming 480p at 60 FPS on one RTX 5090; its model card sets limits on those claims.

  0 upvotes

• Clueso says enterprise retention is nearly 100% as existing investors back $10M round

  The company says fast growth, capital efficiency, its product and customer love brought investors back.

  0 upvotes

• Meta releases Muse Spark Open, an 816B-parameter model built for agents

  Alexandr Wang's team is releasing the weights for coding and long-horizon tasks, with local operation requiring two M5 Ultra Mac Studios or substantial GPU memory.

  1 upvote

• NVIDIA releases a 120B open multimodal model for AI agents

  Nemotron 3.5 Super VL combines text, image, document and video processing; NVIDIA says it completed 10,000 coding tasks up to four times faster in its tests.

  1 upvote

• Cyfendo tests code flaws before developers approve fixes

  Cyfendo founder Bryan Vuong is targeting engineering teams without dedicated security staff; developers review and approve proposed fixes.

  0 upvotes

• Google brings Gemini 4 Argon to Pro subscribers and Search

  The model debuted on September 30th with access limited to cyber defenders; Google opened its Gemini app and AI Mode rollout on October 13th.

  1 upvote

• StepFun publishes 600B model weights with NVIDIA GB300 serving support

  Step 5 Preview has 600 billion total parameters and 27 billion active per token. NVIDIA says its NVFP4 version runs at nearly 8,000 tokens per second per GPU on GB300.

  1 upvote

• Sakana AI builds evidence checks into its open-source AI scientist

  David Ha's third-generation research system assigns coding agents parallel experiments, baselines and ablations, while researchers remain responsible for judging the results.

  0 upvotes

• Bolt Graphics seeks $100 million at a $700 million valuation before Zeus ships

  Reuters reported the proposed raise on October 15th, two days after Bolt said its Zeus test chip had powered on; commercial GPUs are planned for 2028.

  0 upvotes

• Proofpoint agrees to buy Varonis for $6.3B to bolster AI security

  The all-cash deal values Varonis at $49.75 a share and would take the Nasdaq-listed company private if it clears shareholder and regulatory approvals.

  0 upvotes

• TBC says TIME recognized the technology behind its first AI model

  TBC's first product runs on conventional chips; the company says it makes video generation faster and cheaper, claims that remain unverified by independent testing.

  1 upvote

• AstraZeneca will pay insitro $40M upfront, with $1.46B tied to milestones

  The collaboration pairs insitro founder Daphne Koller's biology and machine-learning platform with AstraZeneca's drug-development expertise to pursue MASH and cardiomyopathy treatments.

  1 upvote

• Cua launches shared desktops for teams and AI agents

  Founded by former Microsoft engineer Francesco Bonacci, Cua lets teams connect their own machines, supervise agents on shared desktops and pay for collaboration tools.

  0 upvotes

• MacPaw launches Eney, a Mac assistant with approval controls

  Founder Oleksandr Kosovan is extending MacPaw's Mac-utility business into AI task automation. Eney asks before consequential actions by default, while its AI requests are processed in the cloud.

  1 upvote

• XGRIDS maps Minning Town in 6.2 billion Gaussian splats for film production

  The company says the 23.5-square-kilometer reconstruction combines Lixel L2 Pro and PortalCam capture with an Unreal Engine workflow.

  0 upvotes

• Cellino launches Reservoir, charging $4,995 to bank immune cells

  Nabiha Saklayen's consumer service stores cells below -150 degrees Celsius, while leaving any future medical use to the customer and their physician.

  0 upvotes

• ElevenLabs automates video scores and sound-effect placement in one pass

  The feature pairs original music with effects placed at selected moments, and is available through ElevenCreative, ElevenMusic and ElevenAPI.

  0 upvotes

• QumulusAI turns $71.9M Blackwell contract into production as contracts hit $523M

  QumulusAI CEO Michael Maniscalco, a former Applied Digital CTO, says 7 MW is billing. First-half 2026 revenue was $10.1M.

  1 upvote

• Vercel launches Token Pass at $30 for $80 in model credits

  The subscription bundles GLM 5.3 and Kimi K3 through Vercel AI Gateway, with zero data retention enabled by default and early access capped at 500 Pro teams.

  0 upvotes

• wisepatientAI announces AI record search already listed in January 2025

  CEO and co-founder Arvind Raichur says patients can ask plain-language questions about their records. wisepatientAI described natural-language search in October 2025, and its App Store version history lists the feature in January 2025.

  1 upvote

• Poetiq says its system optimized an LLM router in 12 hours

  The test used simulated company workloads; Poetiq says its custom routers improved cost and accuracy against OpenRouter and TypeSafe's JEV on its benchmark.

  0 upvotes

• Origin secures $20M to move its construction robot beyond drywall

  Co-founded by Amazon-exit veteran Yogesh Ghaturle, Origin will build its OG-1 fleet and pilot the robot with contractors in the U.S. and Europe.

  0 upvotes

• Cognition launches Devin Oncall to investigate and fix production incidents

  The product connects to observability and ticketing tools, investigates alerts and bugs, and can open tested pull requests; Cognition says it already handles thousands of alerts daily.

  0 upvotes

• Katie Zacharia promoted Shakeology through Beachbody's former MLM network

  BODi confirmed she was one of 300,000 independent distributors in 2013; the company ended its multilevel marketing program in 2025.

  0 upvotes

• UpCodes launched Plan Review; Reynolds says estimates draw on users, studies

  Scott Reynolds told RuntimeWire the potential-cost estimates draw on customer and partner accounts, alongside third-party research.

  0 upvotes

• ModelBest's 35B vision model targets phones with 3 GB of memory

  CEO Li Dahai's team says expert preloading makes the open model practical on mobile; the weights still rely on device storage, and the speed figures are company-reported.

  1 upvote

• Better appoints former X CEO Linda Yaccarino to its Better 2.0 board

  Better's first Better 2.0 board appointment comes days after shareholders restored founder Vishal Garg's control of the mortgage lender's board.

  0 upvotes

• Undecimal launches with $18M; Erlich reports mouse mRNA repression

  Yaniv Erlich, in answers relayed by Lily Jeffery, told RuntimeWire that siRNA screen hits produced statistically significant target-mRNA repression in mice after subcutaneous delivery to extrahepatic tissues, but public materials omit key experimental details.

  0 upvotes

• BloomX raises $13M to take its pollination robots into U.S. orchards

  BloomX says its crop-specific systems lift yields by about 20%; founder Thai Sade drew on nearly a decade managing his family's food business.

  0 upvotes

• Sunbound raises $24M and acquires Richter to expand into skilled nursing

  Co-founder Manny Cominsky is pairing revenue software with Richter's billing expertise as Sunbound expands into skilled nursing.

  0 upvotes

• Marklo Raises $5M, Says 45% of 100-Plus Brands Use Product Weekly

  The Shopify-focused startup has opened a free tier and Pro trial after a restricted pilot, founder and CEO Jessica Herrin told RuntimeWire.

  0 upvotes

• FTC investigates Anthropic and OpenAI over AI safety incidents

  The October 14th demands seek incident records, security controls and marketing materials as the agency examines possible consumer-protection violations.

  0 upvotes

• CarbonAI raises $12M to automate commercial real-estate lending

  Co-founders Jeff Saul and Josh Kaufman built the New York startup after selling their previous CRE lending platform to Built; Cowboy Ventures led the funding.

  0 upvotes

• Zesty adds automated allocation to NVIDIA's open-source GPU fractioning

  CEO and co-founder Maxim Melamedov is extending Zesty's cloud-cost thesis to GPU sharing, two days after NVIDIA open-sourced Run:ai.

  0 upvotes

• Meta turns Doja Cat's tour into a 12-minute smart-glasses film

  Directed by Quinn Whitney Wilson, the short premieres October 23rd as Meta pairs a wearable-camera showcase with a Los Angeles retail event.

  0 upvotes

• PointClickCare adds former Appen CEO to board as it expands healthcare AI

  Armughan Ahmad joins the board as Dave Wessinger pursues an AI-first strategy at PointClickCare, founded to keep technology out of clinicians' way.

  0 upvotes

• Fleet adds AI governance to endpoint management as AI moves onto laptops

  CEO Mike McNeil says IT teams should govern AI agents through the device tools they already use; Fleet says inventory is free, while some controls are still upcoming.

  0 upvotes

• Opsin raises $21M to track what AI agents can reach and do

  Business Insider reported that SYN Ventures and Race Capital led the round, which Opsin plans to use to expand sales and engineering.

  0 upvotes

• Otter Health raises $8M to measure care beyond prior authorization

  Keith Florance's new venture pairs real-time clinical reviews with payer-provider feedback; Fika Ventures and Evidenced co-led the seed round.

  0 upvotes

• BAND launches a desktop control room for coding agents

  Arick Goomanovsky's new desktop app connects agent work across projects and environments; BAND raised $17 million from Team8, Hetz Ventures and Sierra Ventures in April.

  0 upvotes

• Discover-Tech ads delivered spyware to Al Jazeera devices, researchers say

  Citizen Lab linked the Israeli adtech company's systems to zero-click Candiru attacks targeting Al Jazeera in 2024 and 2025. The lab says at least 45 devices were targeted; the Financial Times reported at least 45 infections.

  0 upvotes

• Mueon raises $65M to validate modular AI data center systems

  Maverick Silicon led the round, with Intel Capital, Geodesic Alliances Fund, Oregon Venture Fund, M-Ventures, Wellington Capital, IQT and Defined participating. Mueon plans to move Cubelets toward system validation and production readiness.

  0 upvotes

View all Stories