# Loaded Tokens
> The latest AI news in five minutes or less, with a short take on what is worth using, what to keep an eye on, and what you can skip. AI news, learning and product context for humans and agents.

## Scope and evidence
Public-source editorial work. No hands-on tests are claimed. Use means a documented capability fits the stated reader need; Watch means material uncertainty; Ignore means not a fit for the stated need. Prices are dated snapshots, not transaction quotes. Conceptual diagrams and examples are not benchmark results. Draft issues and paid guide bodies are excluded from this public corpus. Source pages and corrections take precedence over cached copies. Corrections: if a fact is wrong, email hello@loadedtokens.com.

## Human entry points
- [News](https://loadedtokens.com/news/)
- [Newsfeed](https://loadedtokens.com/newsfeed/)
- [This week in AI](https://loadedtokens.com/this-week/)
- [Search](https://loadedtokens.com/search/)
- [Learn: five structured tracks](https://loadedtokens.com/learn/)
- [Guides](https://loadedtokens.com/guides/)
- [Best lists](https://loadedtokens.com/best/)
- [Verdict ledger](https://loadedtokens.com/verdicts/)
- [AI for your job](https://loadedtokens.com/for/)
- [Allow the briefing](https://loadedtokens.com/safelist/)
- [Tools and companies](https://loadedtokens.com/directory/)
- [Compare tools](https://loadedtokens.com/tools/compare/)
- [Tool finder](https://loadedtokens.com/tools/quiz/)
- [Pricing changelogs](https://loadedtokens.com/tools/pricing/)
- [Glossary: 54 sourced definitions](https://loadedtokens.com/learn/glossary/)
- [Topics: connected news, tools and learning](https://loadedtokens.com/topics/)
- [Agent guide and field conventions](https://loadedtokens.com/for-agents/)

## Machine-readable exports
- [Full public Markdown corpus](https://loadedtokens.com/llms-full.txt)
- [Learn JSON](https://loadedtokens.com/learn.json): tracks, goals, diagrams, sources and full Markdown
- [Glossary JSON](https://loadedtokens.com/glossary.json): definitions, evidence and deep links
- [Tool directory JSON](https://loadedtokens.com/directory.json): public evidence, dated pricing, audience and verdicts
- [Company JSON](https://loadedtokens.com/companies.json): profiles, pricing and sources
- [Topics JSON](https://loadedtokens.com/topics.json): categories and public-content counts
- [Issues JSON](https://loadedtokens.com/issues.json)
- [Newsfeed JSON](https://loadedtokens.com/newsfeed.json)
- [Guides JSON](https://loadedtokens.com/guides.json)
- [Verdicts JSON](https://loadedtokens.com/verdicts.json)
- [This week JSON](https://loadedtokens.com/this-week.json)
- [Newsfeed JSON Feed](https://loadedtokens.com/newsfeed/feed.json)
- [Newsfeed RSS](https://loadedtokens.com/newsfeed/rss.xml)
- [Search index](https://loadedtokens.com/search.json)
- [Compare JSON](https://loadedtokens.com/tools/compare.json)
- [Tool finder JSON](https://loadedtokens.com/tools/quiz.json)
- [Alternatives JSON](https://loadedtokens.com/alternatives.json)
- [Pricing changelog JSON](https://loadedtokens.com/pricing.json)
- [Tools API](https://loadedtokens.com/api/tools.json): filter with `?category=` or `/api/tools/<category>.json`
- [Open source hub](https://loadedtokens.com/categories/open-source/)
- [Resources](https://loadedtokens.com/resources/)
- [Newsfeed Atom](https://loadedtokens.com/newsfeed/atom.xml)
- [JSON Feed](https://loadedtokens.com/feed.json)
- [RSS](https://loadedtokens.com/rss.xml)

## Newsfeed
- [Two Chinese chip suppliers set for Hong Kong trading debut](https://loadedtokens.com/newsfeed/two-chinese-chip-suppliers-set-for-hong-kong-tra-6d0609/): 2026-11-09T23:25:31.000Z. watch. Bloomberg reports that Suzhou Denglin Technology and Jiangsu Xinde Semiconductor Technology were due to start trading in Hong Kong on Tuesday, after their IPOs raised a combined $301 million. According to Bloomberg, the listings arrive while tech stocks are swinging sharply, so the debuts will test how much investors want chip shares. Source: https://www.bloomberg.com/news/articles/2026-11-09/china-chipmakers-to-debut-in-hk-in-test-of-tech-stock-demand
- [Semafor: Congress turns its attention to Nvidia's CEO after a spy scandal](https://loadedtokens.com/newsfeed/semafor-congress-turns-its-attention-to-nvidia-s-f28431/): 2026-11-09T23:11:24.000Z. watch. Semafor reports that lawmakers are looking at Nvidia's chief executive after a spy scandal. According to the outlet, they are considering tighter export controls that would restrict the company's chip sales to Beijing, along with possible investigations. The excerpt does not describe the scandal itself. Source: https://www.semafor.com/article/11/09/2026/nvidia-ceo-is-in-congress-sights-after-spy-scandal
- [Nvidia shares drop after board probes Jensen Huang's ties to suspected Chinese spy](https://loadedtokens.com/newsfeed/nvidia-shares-drop-after-board-probes-jensen-hua-bda301/): 2026-11-09T22:23:42.000Z. watch. TechCrunch reports that Nvidia's stock fell more than 4% after its board began investigating Jensen Huang's relationship with a person suspected of spying for China. A securities law expert quoted by the outlet warned that Nvidia's earlier claim that the matter was "normal" could prove costly if shareholders sue. Source: https://techcrunch.com/2026/11/09/the-race-is-on-to-sue-nvidia/
- [Bloomberg: Lawmakers seek answers on Nvidia after report on Jensen Huang and alleged Chinese spy](https://loadedtokens.com/newsfeed/bloomberg-lawmakers-seek-answers-on-nvidia-after-c35d26/): 2026-11-09T21:57:43.000Z. watch. Bloomberg reports that Nvidia CEO Jensen Huang had an affair with a person alleged to be a Chinese spy. According to the outlet, members of both parties in Congress are now pressing for more information. They want to gauge what the matter could mean for national security. The excerpt describes allegations and early demands, not findings or any action taken. Source: https://www.bloomberg.com/news/articles/2026-11-09/us-lawmakers-demand-details-on-nvidia-amid-jensen-huang-scandal
- [Bloomberg: Nvidia's top lawyer stepped into US chip-smuggling probe of Megaspeed](https://loadedtokens.com/newsfeed/bloomberg-nvidia-s-top-lawyer-stepped-into-us-ch-63c359/): 2026-11-09T21:28:26.000Z. watch. Bloomberg reports that Nvidia's general counsel got involved in a US government investigation into suspected AI chip smuggling by Singapore-based Megaspeed International. The lawyer reportedly asked whether the Trump administration had a hand in detaining a Megaspeed executive. Bloomberg says that executive is now central to a scandal that threatens Jensen Huang. Source: https://www.bloomberg.com/news/articles/2026-11-09/nvidia-intervened-in-us-probe-of-megaspeed-chip-smuggling
- [LangChain says agent context engineering starts with better retrieval](https://loadedtokens.com/newsfeed/langchain-says-agent-context-engineering-starts-2b4301/): 2026-11-09T21:20:38.000Z. watch. LangChain has a new post arguing that agents burn time and context on clumsy searches, oversized results and outdated data. The company describes the retrieval techniques its Deep Agents harness uses to give agents the right context. Builders working on agent memory may find it a useful practical reference. Source: https://www.langchain.com/blog/context-engineering-starts-with-retrieval
- [Google pushes new AI chips as it tries to take on Nvidia](https://loadedtokens.com/newsfeed/google-pushes-new-ai-chips-as-it-tries-to-take-o-387e80/): 2026-11-09T21:12:56.000Z. watch. Bloomberg reports that Google is releasing a new set of AI chips to build on its rise as a competitor to Nvidia. The excerpt is a short video segment featuring Bloomberg Intelligence analyst Mandeep Singh. It gives no chip names, specs, pricing or availability, so the business details still need confirming. Source: https://www.bloomberg.com/news/videos/2026-11-09/google-rolls-out-new-ai-chips-to-challenge-nvidia-video
- [Fireworks AI tells customers of a second breach within weeks, Bloomberg reports](https://loadedtokens.com/newsfeed/fireworks-ai-tells-customers-of-a-second-breach-0ac84c/): 2026-11-09T20:46:58.000Z. use. Bloomberg reports that Fireworks AI, an Nvidia-backed startup that runs AI models for other companies, told customers on Monday that attackers had accessed their data. According to Bloomberg, this is the company's second known hack in a few weeks. The excerpt doesn't say what data was exposed or how many customers were affected. Source: https://www.bloomberg.com/news/articles/2026-11-09/nvidia-backed-fireworks-ai-notifies-customers-of-its-second-hack-in-weeks
- [Guardian: Nvidia board faces hard choice over Jensen Huang's future](https://loadedtokens.com/newsfeed/guardian-nvidia-board-faces-hard-choice-over-jen-1dd2cf/): 2026-11-09T20:27:07.000Z. watch. The Guardian reports that Nvidia's board must make a difficult decision about Jensen Huang after the co-founder had an affair with an alleged Chinese spy. Huang has led the chipmaker since he co-founded it in 1993, and the outlet notes his children hold leadership roles there. The piece asks whether he could be pushed out. The excerpt does not report a board decision. Source: https://www.theguardian.com/technology/2026/nov/09/jensen-huang-image-nvidia
- [Bedrock Managed Knowledge Base adds single-shot and retrieval-only modes alongside agentic retrieval](https://loadedtokens.com/newsfeed/bedrock-managed-knowledge-base-adds-single-shot-b09ddd/): 2026-11-09T20:13:18.000Z. use. AWS says Amazon Bedrock Managed Knowledge Base now handles single-shot RAG and retrieval-only requests through AgenticRetrieveStream, and agentic retrieval is still available. AWS says a parameter called maxAgentIteration controls how much retrieval effort each request gets. Its walkthrough uses a made-up internal banking policy assistant as the example. Source: https://aws.amazon.com/blogs/machine-learning/single-shot-and-agentic-retrieval-with-amazon-bedrock-managed-knowledge-base/
- [NVIDIA's Nemotron 3.5 Super VL arrives on Amazon SageMaker JumpStart](https://loadedtokens.com/newsfeed/nvidia-s-nemotron-3-5-super-vl-arrives-on-amazon-1cff12/): 2026-11-09T20:01:20.000Z. use. AWS says NVIDIA's Nemotron 3.5 Super VL, a 120B-parameter vision-language model, can now be deployed from SageMaker JumpStart. Teams can choose an NVFP4 or BF16 variant and run inference on text, images, and video. AWS is positioning it for document understanding, computer-use agents, and other multimodal agent work. Source: https://aws.amazon.com/blogs/machine-learning/nvidia-nemotron-3-5-super-vl-is-now-available-in-amazon-sagemaker-jumpstart/
- [Reindeer brings its Gemini-based enterprise agents to Google Cloud Marketplace](https://loadedtokens.com/newsfeed/reindeer-brings-its-gemini-based-enterprise-agen-63d354/): 2026-11-09T19:42:00.000Z. watch. In a press release, Reindeer says its enterprise AI agents can now be bought by Google Cloud customers through the Google Cloud Marketplace. The company describes the agents as built on Gemini models and deployed on Google Cloud, and it calls the move a wider partnership aimed at core business operations. This is a vendor announcement, and the excerpt gives no customer results or pricing. Source: https://www.prnewswire.com/news-releases/reindeer-launches-on-google-cloud-marketplace-expanding-collaboration-with-google-cloud-to-help-enterprises-move-agentic-ai-into-core-operations-302933757.html
- [US stocks recover from AI sell-off even as Nvidia slips, FT reports](https://loadedtokens.com/newsfeed/us-stocks-recover-from-ai-sell-off-even-as-nvidi-bab4f9/): 2026-11-09T19:24:57.000Z. watch. The Financial Times reports that US stock indices bounced back after a sell-off tied to AI. Lower oil prices helped the broader market look past a decline in Nvidia shares. The excerpt does not give figures or explain what started the AI rout. That makes this a read on investor mood, not news about the technology. Source: https://www.ft.com/content/b8644327-86e7-4d1c-a71f-3fb1d552723b?syn-25a6b1a6=1
- [Claude Code v2.1.321 makes Claude Fable 5.2 the default Fable model and fixes an MCP OAuth token leak](https://loadedtokens.com/newsfeed/claude-code-v2-1-321-makes-claude-fable-5-2-the-ace31f/): 2026-11-09T19:22:30.000Z. use. Anthropic's Claude Code release notes say Claude Fable 5.2 is now the default Fable model, with a 1M-token context window. It is priced at $8 input and $40 output per million tokens, and cache reads cost $0.20. The update also fixes a bug where MCP OAuth could pass refresh tokens to a different host after a server redirect. Source: https://github.com/anthropics/claude-code/releases/tag/v2.1.321
- [Apple adds a $49.99 option to buy more Apple Intelligence usage in Logic Pro](https://loadedtokens.com/newsfeed/apple-adds-a-49-99-option-to-buy-more-apple-inte-a311ac/): 2026-11-09T19:09:58.000Z. watch. The Verge reports that Logic Pro now has a "Buy More Usage" option, a one-time $49.99 purchase. It extends access to Apple Intelligence features that carry usage limits. Apple first described those limits in a support article in early September. According to the report, 9to5Mac spotted the change, and Logic Pro includes a built-in AI assistant that can help with tasks such as starting a new song. Source: https://www.theverge.com/tech/1030084/apple-intelligence-usage-packs-ai-features-creator-studio-logic-pro
- [OpenAI marks ChatGPT conversation error incident as resolved](https://loadedtokens.com/newsfeed/openai-marks-chatgpt-conversation-error-incident-f5ba1e/): 2026-11-09T19:08:31.461Z. ignore. OpenAI's status page lists an incident with elevated errors in ChatGPT conversations, now marked resolved. The company says every affected service is back to normal and the Conversations component shows as operational. The post does not give a cause, duration, or how many users were hit. Source: https://status.openai.com//incidents/01M70RDA2GWD2WXE58ANY0WV5R
- [Bloomberg: Anthropic releases new Mythos and Fable models as its IPO approaches](https://loadedtokens.com/newsfeed/bloomberg-anthropic-releases-new-mythos-and-fabl-6203cf/): 2026-11-09T19:00:20.000Z. watch. Bloomberg reports that Anthropic has released updated versions of its Mythos and Fable AI models. According to the outlet, the launch is meant to strengthen Anthropic's standing as a leading AI developer just before the company goes public. The excerpt doesn't describe what's new in the models, how they perform, or when they'll be available. Source: https://www.bloomberg.com/news/articles/2026-11-09/anthropic-unveils-new-fable-mythos-ai-models-ahead-of-ipo
- [Anthropic rolls out Claude Fable 5.2 and Mythos 5.2, The Verge reports](https://loadedtokens.com/newsfeed/anthropic-rolls-out-claude-fable-5-2-and-mythos-719be9/): 2026-11-09T19:00:00.000Z. watch. The Verge reports that Anthropic has released two models, Claude Fable 5.2 and Claude Mythos 5.2, and is positioning them for hard technical work. Anthropic says both improve on the 5.1 versions in coding, reasoning, and vision, and also produce output faster. These are the company's own claims. The report does not include any independent testing. Source: https://www.theverge.com/ai-artificial-intelligence/1030005/anthropic-claude-fable-mythos-5-2
- [Ericsson's new CEO says AI devices will drive demand for next-gen networks](https://loadedtokens.com/newsfeed/ericsson-s-new-ceo-says-ai-devices-will-drive-de-87b165/): 2026-11-09T15:14:26.000Z. watch. Bloomberg reports that Ericsson's new chief executive expects artificial intelligence to boost demand for next-generation telecom networks. According to the report, the reasoning is that AI-connected gadgets such as smart glasses will need more data. For now this is an executive's outlook and does not include figures or a timeline. Source: https://www.bloomberg.com/news/articles/2026-11-09/new-ericsson-ceo-sees-ai-fueling-next-generation-network-demand
- [Digital Realty brings its liquid cooling lab program to São Paulo](https://loadedtokens.com/newsfeed/digital-realty-brings-its-liquid-cooling-lab-pro-b5f73c/): 2026-11-09T15:13:16.000Z. watch. Data Center Dynamics reports that Digital Realty has opened a liquid cooling lab in São Paulo, Brazil. According to the outlet, it is the company's first DRIL lab in Latin America. It comes after a London lab that opened earlier this year. Source: https://www.datacenterdynamics.com/en/news/digital-realty-launches-liquid-cooling-lab-in-s%C3%A3o-paulo-brazil/
- [Claude's memory gets a dreaming pass](https://loadedtokens.com/newsfeed/anthropic-agent-memory/): 2026-11-06T12:00:00-06:00 filed. watch. Anthropic's engineering post describes how Claude chooses what to remember, including a dreaming pass that organizes memory and corrects mistakes. It explains the product. It does not hand you a new setting. You cannot schedule the dream from the blog post. Source: https://www.anthropic.com/research/components-of-agent-memory
- [Can the agent read a messy lab bench?](https://loadedtokens.com/newsfeed/biobasicsbench/): 2026-11-06T12:00:00-06:00 filed. watch. OpenAI and Ligo published BioBasicsBench, 100 tasks about raw wet-lab evidence, including the times the data cannot support a conclusion. It scores agents. It does not hand you a lab partner. Source: https://openai.com/index/measuring-how-ai-agents-handle-wet-lab-science
- [You can phone a Cursor agent now](https://loadedtokens.com/newsfeed/cursor-voice-chat/): 2026-11-06T12:00:00-06:00 filed. watch. You can talk to a Cursor agent while it works, on the desktop app, the web, and iOS. The call runs on Grok Voice. Dictation is still the quieter cousin: it only types into the box. Source: https://cursor.com/changelog/voice-chat
- [OpenAI hired a bug hunter for its own house](https://loadedtokens.com/newsfeed/openai-hacktron/): 2026-11-06T12:00:00-06:00 filed. watch. OpenAI will work with Hacktron to hunt bugs in OpenAI's own systems, especially where those systems touch third parties. Hacktron also gets broader Daybreak Red access. This is a security program. There is no switch in your settings. Source: https://openai.com/index/partnering-with-hacktron
- [Perplexity's changelog mentions a local Mac app](https://loadedtokens.com/newsfeed/perplexity-local-mac/): 2026-11-06T12:00:00-06:00 filed. watch. A November 6 changelog note lists a fully local Computer on Mac, plus a voice mode next to Teammates. The note does not spell out plan limits. Read it before you promise either feature in Slack. Source: https://www.perplexity.ai/changelog/teammates-fully-local-computer-on-mac-and-voice-mode
- [AWS will answer the web in one tool call](https://loadedtokens.com/newsfeed/aws-web-answer/): 2026-11-05T12:00:00-06:00 filed. watch. Web Answer on Bedrock AgentCore is generally available and returns a cited answer over MCP. It is billed per call and, for now, only in US East (N. Virginia). If your stack lives somewhere else, this is a postcard. Source: https://aws.amazon.com/about-aws/whats-new/2026/11/web-answer-amazon-bedrock/
- [A leaked Claude key can be fenced in](https://loadedtokens.com/newsfeed/claude-ip-allowlist/): 2026-11-05T12:00:00-06:00 filed. use. Claude Console organizations can set an IP allowlist, in beta, with up to 1,000 addresses or ranges. Requests from anywhere else are refused, which limits what a leaked key can reach. It does not replace rotating the key. Change the locks too. Source: https://platform.claude.com/docs/en/release-notes/overview
- [Claude Enterprise can sign someone out everywhere](https://loadedtokens.com/newsfeed/claude-sign-out/): 2026-11-05T12:00:00-06:00 filed. watch. Claude Enterprise admins can sign a member out of Claude everywhere, in beta, through the Admin API. The person stays in the organization and can sign in again. It is an off switch for sessions. Their name stays on the company list. Source: https://platform.claude.com/docs/en/release-notes/overview
- [Cursor will build the app, then wait for you to hit publish](https://loadedtokens.com/newsfeed/cursor-live-apps/): 2026-11-05T12:00:00-06:00 filed. watch. A cloud agent can scaffold an app, add Postgres and Cursor sign-in, and deploy it after security checks. You need a paid plan with cloud agents. Nothing goes live until you explicitly publish. Source: https://cursor.com/changelog/build-live-apps
- [Baidu shipped a base model, not a chatbot you can ping](https://loadedtokens.com/newsfeed/ernie-5-5-flash-base/): 2026-11-05T12:00:00-06:00 filed. watch. ERNIE 5.5 Flash Base is a 280-billion-parameter mixture-of-experts checkpoint, 14 billion active, stopped before instruction tuning. The license does not cover the EU, the UK, or South Korea. Check the map before the model card gets you excited. Source: https://ernie.baidu.com/release/ernie-5-5-flash-base/
- [Fifteen seconds of video, priced like a coffee add-on](https://loadedtokens.com/newsfeed/gemini-omni-1-2-flash/): 2026-11-05T12:00:00-06:00 filed. use. Gemini Omni 1.2 Flash makes native clips up to 15 seconds. Google priced 720p at $0.07 a second, 30% under Omni 1.1. It is generally available on the Gemini API and Google Cloud, and it stamps a SynthID watermark. Source: https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemini-omni-1-2-flash/
- [GPT-6 Cyber is not your Monday chatbot](https://loadedtokens.com/newsfeed/gpt-6-cyber/): 2026-11-05T12:00:00-06:00 filed. watch. OpenAI's new cybersecurity model is rated Critical under its Preparedness Framework and starts with approved Daybreak Red customers. It finds and fixes vulnerabilities inside an authorized scope. The all-hands email is still your job. Source: https://openai.com/index/introducing-gpt-6-cyber
- [Inkling 2 Small is big, closed, and on Tinker first](https://loadedtokens.com/newsfeed/inkling-2-small/): 2026-11-05T12:00:00-06:00 filed. watch. Thinking Machines' new model has 594 billion parameters, 17 billion active, and a score of 31 on the Artificial Analysis Intelligence Index. Weights stay closed until outside safety testing. Apache 2.0 is the plan, not today's license. Source: https://thinkingmachines.ai/news/inkling-2-small/
- [NTT DATA orders 900 megawatts of fuel cells](https://loadedtokens.com/newsfeed/ntt-bloom-fuel-cells/): 2026-11-05T12:00:00-06:00 filed. watch. NTT DATA said on November 5 it will deploy 900 megawatts of Bloom Energy fuel cells across data center projects, with first shipments planned for 2027. The release calls this Bloom's largest international project, and says the cells make electricity without combustion. Bloom flags the megawatts, the start date, and any expansion as forward-looking. Nothing in the release is a site you can tour. Source: https://www.nttdata.com/en-us/news/2026/ntt-data-announces-agreement-with-bloom-energy
- [OpenAI wrote a note to the towns it builds in](https://loadedtokens.com/newsfeed/openai-infrastructure/): 2026-11-05T12:00:00-06:00 filed. watch. OpenAI published community commitments for the places it builds infrastructure, plus policy proposals it says would spread the benefits. Read it as a company position. It will not pick your next API. Source: https://openai.com/index/setting-new-standard-responsible-ai-infrastructure/
- [A speech model small enough to live on the device](https://loadedtokens.com/newsfeed/phonon-tts/): 2026-11-05T12:00:00-06:00 filed. watch. Gradium's Phonon runs without a server, in about 65 MB and 75 MB sizes, across five languages. The English 31-million-parameter model is Apache 2.0. The other models are non-commercial. Local does not mean "sell it in your app." Source: https://huggingface.co/blog/phonon-tts
- [CUDA 13.5 shows up for Vera Rubin](https://loadedtokens.com/newsfeed/cuda-13-5/): 2026-11-04T12:00:00-06:00 filed. use. NVIDIA's toolkit release adds general-availability support for the Vera Rubin architecture and extends Windows on Arm to DGX Station. If you already build on CUDA, this is the upgrade note, not a new app for your lunch break. Source: https://developer.nvidia.com/blog/nvidia-cuda-13-5-brings-nvidia-vera-rubin-ga-support/
- [Google published a report card for health AI](https://loadedtokens.com/newsfeed/google-health-eval/): 2026-11-04T12:00:00-06:00 filed. watch. Google published a health-AI evaluation framework written with outside clinicians. It looks at accuracy, safety, and usefulness, using benchmarks, autoraters, and expert review. The post shares the method. It does not publish a score for a model you can buy. Source: https://blog.google/innovation-and-ai/technology/health/evaluating-health-ai-at-scale/
- [ChatGPT images learned how to loop](https://loadedtokens.com/newsfeed/images-2-6/): 2026-11-04T12:00:00-06:00 filed. use. Images 2.6 is in ChatGPT, Codex, and the API, with GIF output and generation OpenAI says is 50% faster than before. It is available on every ChatGPT subscription tier. Your mascot can wave. A person still has to approve it. Source: https://openai.com/index/introducing-images-2-6/
- [A lab pipeline wrote 948 Linux exploits](https://loadedtokens.com/newsfeed/openai-linux-lpe/): 2026-11-04T12:00:00-06:00 filed. watch. OpenAI says a research pipeline produced working local privilege-escalation exploits for 948 Linux kernel issues and drafted patches. The post puts that run at $1,213,430 in API credits. It is a controlled lab result. You cannot point this at the office server. Source: https://openai.com/index/a-thousand-fewer-paths-to-root/
- [OpenAI wants the Git tools already inside Codex](https://loadedtokens.com/newsfeed/openai-pierre/): 2026-11-04T12:00:00-06:00 filed. watch. OpenAI says it will acquire Pierre Computer, maker of Code Storage and the Trees and Diffs libraries already used in Codex. The deal has not closed. The companies stay separate until it does. Your repo does not move houses today. Source: https://openai.com/index/openai-to-acquire-pierre/
- [Perplexity's teammates are a Max-plan hire](https://loadedtokens.com/newsfeed/perplexity-teammates/): 2026-11-04T12:00:00-06:00 filed. watch. Teammates are personal agents for work that runs past one prompt, and they keep project memory across runs. Perplexity is launching them on the web for Max subscribers, with iOS listed as later. Not on Max? You are watching the trailer. Source: https://www.perplexity.ai/hub/blog/teammates-are-agents-for-complex-ongoing-work
- [Some Claude text now carries a watermark](https://loadedtokens.com/newsfeed/claude-text-watermark/): 2026-11-03T12:00:00-06:00 filed. watch. As of November 3, text from Claude Mythos 5, Opus 4.7, and Sonnet 4.6 carries Anthropic's text watermark on the platforms named in its marking guide. On Sonnet 4.6, a non-default temperature, top_p, or top_k may skip the mark. Source: https://platform.claude.com/docs/en/release-notes/overview
- [Long documents still lose the plot](https://loadedtokens.com/newsfeed/statebench/): 2026-11-03T12:00:00-06:00 filed. watch. Ai2's StateBench checks whether a model can track changing facts across a long document, at four levels of difficulty. Frontier models handle the easy levels and break down on the hardest long ones. It is a benchmark. It will not file your contracts. Source: https://allenai.org/blog/statebench
- [Claude will hold the thread, for five minutes](https://loadedtokens.com/newsfeed/claude-message-threads/): 2026-11-02T12:00:00-06:00 filed. watch. Message threads are in public beta on the Claude Platform. After the first request, later turns send only the new messages and the previous response id. A thread expires five minutes after the last request. Anthropic says token use stays the same. Source: https://platform.claude.com/docs/en/release-notes/overview
- [A receipt that says the model was not swapped](https://loadedtokens.com/newsfeed/flashprove/): 2026-11-02T12:00:00-06:00 filed. watch. Anthropic open-sourced FlashProve, a research prototype that proves a reply came from a committed set of weights. It is Apache 2.0 research code, not a Claude setting. Proving costs many times a normal answer. Nice receipt. Expensive printer. Source: https://www.anthropic.com/research/provable-inference
- [LifeVal grades the errands, not the trivia night](https://loadedtokens.com/newsfeed/lifeval/): 2026-11-02T12:00:00-06:00 filed. watch. LifeVal is OpenAI's evaluation of everyday tasks, from childcare and budgets to government paperwork. The post explains how the task list was built. It is a measurement project. You cannot switch it on in ChatGPT. Source: https://openai.com/index/lifeval/
- [OpenAI raised the bounty for the bugs it missed](https://loadedtokens.com/newsfeed/openai-bug-bounty/): 2026-11-02T12:00:00-06:00 filed. watch. OpenAI raised Security Bug Bounty payouts on November 2. A critical finding in a core product such as ChatGPT, Codex, or the API now pays up to $30,000. Exceptional cases can pay more once bonuses are included. This is a researcher program, not a product setting. Source: https://openai.com/index/rewarding-defenders-with-bigger-bug-bounties/
- [OpenAI's campuses would turn down the power](https://loadedtokens.com/newsfeed/openai-grid-protectors/): 2026-11-02T12:00:00-06:00 filed. watch. On November 2, OpenAI launched Grid Protectors, a plan for future compute campuses to cut electricity use when the local grid is strained. The Effingham County agreement with Georgia Power includes up to 1,000 megawatts of flexible demand response. OpenAI compares that with the average demand of roughly 700,000 households. The campus is still in development. A pledge is not a building with the lights on. Source: https://openai.com/index/grid-protectors-program
- [GPT-Live can place the phone call](https://loadedtokens.com/newsfeed/openai-live-sip/): 2026-11-02T12:00:00-06:00 filed. watch. Outbound SIP calling is now available to every organization on the Live API. You connect GPT-Live to a number through your own SIP provider. The changelog is the availability note. The price sits on a separate pricing page. You still bring the phone plan. Source: https://developers.openai.com/api/docs/changelog
- [A 4B model that clicks the screen and also codes](https://loadedtokens.com/newsfeed/mage-vl-cowork/): 2026-11-01T12:00:00-06:00 filed. watch. Microsoft published Mage-VL-Cowork, a 4-billion-parameter agent that mixes on-screen actions with a terminal, under Apache 2.0. It reports 66.0% on OSWorld-Verified with a 50-step budget. That figure is theirs. We did not sit it at a laptop. Source: https://microsoft.github.io/Mage/cowork/
- [A filing, not a tour: Powerlattice and $86 million](https://loadedtokens.com/newsfeed/powerlattice-sec-filing/): 2026-10-30T12:00:00-06:00 filed. watch. GeekWire says an SEC filing shows Powerlattice raised $86 million and is seeking $111 million, with 18 investors so far. The filing does not name the round, and the company declined to comment. Powerlattice builds a power-delivery chiplet and says a shorter path to the processor can cut energy use by more than half. That cut is the company's claim. We did not open the filing or run the chip. Source: https://www.geekwire.com/2026/filing-chip-startup-powerlattice-raises-86m-to-ease-ais-power-crunch/
- [FuriosaAI raised $620 million for inference chips](https://loadedtokens.com/newsfeed/furiosaai-620m/): 2026-10-28T12:00:00-06:00 filed. watch. SiliconANGLE reported on October 28 that FuriosaAI raised $620 million to scale its inference chips. The company put the valuation at $2.9 billion, and South Korea's National Growth Fund put in $280 million of the round. The article says the chips use less than half the power of Nvidia's GPUs. That comparison is not a test we ran. A bigger round is not a bigger power line. Source: https://siliconangle.com/2026/10/28/energy-efficient-inference-chip-developer-furiosaai-raises-620m-to-solve-the-data-center-power-bottleneck/
- [Gemini 4 Carbon is announced. The door is still locked.](https://loadedtokens.com/newsfeed/gemini-4-carbon/): 2026-10-28T12:00:00-06:00 filed. watch. Google calls Carbon its strongest model and is starting with a subset of cyber defenders, then trusted partners in the Fairwind program. Introductory API pricing is $2 in and $10 out per million tokens. A general release is not dated. Source: https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/
- [Genie 4 builds a world, then hands you a waitlist](https://loadedtokens.com/newsfeed/genie-4/): 2026-10-27T12:00:00-06:00 filed. ignore. DeepMind's Genie 4 generates interactive worlds with native audio. The rollout is a research preview for trusted testers, plus a waitlist. You cannot put this on a work project this week. The trailer is not the product. Source: https://deepmind.google/blog/genie-4-a-new-generation-of-world-simulation/
- [GPT-6.1 Terra is the cheaper workhorse](https://loadedtokens.com/newsfeed/gpt-6-1-terra/): 2026-10-26T12:00:00-06:00 filed. use. OpenAI's fast workhorse is in the API at $1 per million input tokens and $5 per million output. OpenAI says that is 50% below GPT-5.6 Terra and GPT-6.1 Sol. Paid ChatGPT Work and Codex plans can use it too. Source: https://openai.com/index/introducing-gpt-6-1-terra

## This week (2026-11-09 to 2026-11-15)
- [Anthropic rolls out Claude Fable 5.2 and Mythos 5.2, The Verge reports](https://loadedtokens.com/newsfeed/anthropic-rolls-out-claude-fable-5-2-and-mythos-719be9/)
- [Apple adds a $49.99 option to buy more Apple Intelligence usage in Logic Pro](https://loadedtokens.com/newsfeed/apple-adds-a-49-99-option-to-buy-more-apple-inte-a311ac/)
- [Bedrock Managed Knowledge Base adds single-shot and retrieval-only modes alongside agentic retrieval](https://loadedtokens.com/newsfeed/bedrock-managed-knowledge-base-adds-single-shot-b09ddd/)
- [Bloomberg: Anthropic releases new Mythos and Fable models as its IPO approaches](https://loadedtokens.com/newsfeed/bloomberg-anthropic-releases-new-mythos-and-fabl-6203cf/)
- [Bloomberg: Lawmakers seek answers on Nvidia after report on Jensen Huang and alleged Chinese spy](https://loadedtokens.com/newsfeed/bloomberg-lawmakers-seek-answers-on-nvidia-after-c35d26/)
- [Bloomberg: Nvidia's top lawyer stepped into US chip-smuggling probe of Megaspeed](https://loadedtokens.com/newsfeed/bloomberg-nvidia-s-top-lawyer-stepped-into-us-ch-63c359/)
- [Claude Code v2.1.321 makes Claude Fable 5.2 the default Fable model and fixes an MCP OAuth token leak](https://loadedtokens.com/newsfeed/claude-code-v2-1-321-makes-claude-fable-5-2-the-ace31f/)
- [Digital Realty brings its liquid cooling lab program to São Paulo](https://loadedtokens.com/newsfeed/digital-realty-brings-its-liquid-cooling-lab-pro-b5f73c/)
- [Ericsson's new CEO says AI devices will drive demand for next-gen networks](https://loadedtokens.com/newsfeed/ericsson-s-new-ceo-says-ai-devices-will-drive-de-87b165/)
- [Fireworks AI tells customers of a second breach within weeks, Bloomberg reports](https://loadedtokens.com/newsfeed/fireworks-ai-tells-customers-of-a-second-breach-0ac84c/)
- [Google pushes new AI chips as it tries to take on Nvidia](https://loadedtokens.com/newsfeed/google-pushes-new-ai-chips-as-it-tries-to-take-o-387e80/)
- [Guardian: Nvidia board faces hard choice over Jensen Huang's future](https://loadedtokens.com/newsfeed/guardian-nvidia-board-faces-hard-choice-over-jen-1dd2cf/)
- [LangChain says agent context engineering starts with better retrieval](https://loadedtokens.com/newsfeed/langchain-says-agent-context-engineering-starts-2b4301/)
- [Nvidia shares drop after board probes Jensen Huang's ties to suspected Chinese spy](https://loadedtokens.com/newsfeed/nvidia-shares-drop-after-board-probes-jensen-hua-bda301/)
- [NVIDIA's Nemotron 3.5 Super VL arrives on Amazon SageMaker JumpStart](https://loadedtokens.com/newsfeed/nvidia-s-nemotron-3-5-super-vl-arrives-on-amazon-1cff12/)
- [OpenAI marks ChatGPT conversation error incident as resolved](https://loadedtokens.com/newsfeed/openai-marks-chatgpt-conversation-error-incident-f5ba1e/)
- [Reindeer brings its Gemini-based enterprise agents to Google Cloud Marketplace](https://loadedtokens.com/newsfeed/reindeer-brings-its-gemini-based-enterprise-agen-63d354/)
- [Semafor: Congress turns its attention to Nvidia's CEO after a spy scandal](https://loadedtokens.com/newsfeed/semafor-congress-turns-its-attention-to-nvidia-s-f28431/)
- [Two Chinese chip suppliers set for Hong Kong trading debut](https://loadedtokens.com/newsfeed/two-chinese-chip-suppliers-set-for-hong-kong-tra-6d0609/)
- [US stocks recover from AI sell-off even as Nvidia slips, FT reports](https://loadedtokens.com/newsfeed/us-stocks-recover-from-ai-sell-off-even-as-nvidi-bab4f9/)

## Learn
- [AI is a prediction engine, not a tiny coworker](https://loadedtokens.com/learn/what-a-language-model-does/): Understand the difference between producing plausible language and knowing that a claim is true.
- [Tokens, context and memory: three different tabs on the bill](https://loadedtokens.com/learn/tokens-context-and-memory/): Decode the units behind model limits without pretending a token is always a word.
- [Choose a model like a vehicle, not a sports team](https://loadedtokens.com/learn/choose-a-model-by-the-job/): A practical scorecard for comparing capability, latency, data rules and cost.
- [Open weights are not a free lunch with a GPU garnish](https://loadedtokens.com/learn/open-weights-versus-hosted/): Separate access to model files from licenses, hosting costs and operational ownership.
- [An agent is a loop. Your business still needs a brake.](https://loadedtokens.com/learn/agents-versus-workflows/): Know when a fixed workflow beats a model choosing its own next action.
- [MCP is the adapter. Permission is still the socket.](https://loadedtokens.com/learn/mcp-and-tool-permissions/): Understand tool interoperability without mistaking a connection for a safety guarantee.
- [RAG gives your model a reading list, not a truth serum](https://loadedtokens.com/learn/rag-without-the-magic/): How retrieval-augmented generation connects evidence to an answer—and where it can still fail.
- [Your AI demo needs a report card and a receipt](https://loadedtokens.com/learn/evals-and-cost-per-success/): Evaluate quality and total cost per acceptable outcome instead of celebrating one lucky output.
- [Stop writing magic spells. Write a useful brief.](https://loadedtokens.com/learn/prompting-with-a-brief/): A reusable prompt structure built around task, context, constraints and an inspectable output.
- [The workplace AI checklist: helpful, not accidentally public](https://loadedtokens.com/learn/safe-workplace-ai/): Choose low-risk workflows, protect data and measure real effort saved without outsourcing accountability.

## Companies
- [Alibaba Cloud](https://loadedtokens.com/companies/alibaba/): Publisher of the Qwen open-weight model family.
- [Anthropic](https://loadedtokens.com/companies/anthropic/): The company behind Claude and Claude Code.
- [AssemblyAI](https://loadedtokens.com/companies/assemblyai/): Speech-to-text and a newer text-to-speech API.
- [AWS](https://loadedtokens.com/companies/aws/): Amazon Bedrock model services and Amazon Q Developer coding assistance.
- [Black Forest Labs](https://loadedtokens.com/companies/black-forest-labs/): Maker of the FLUX image models.
- [Cartesia](https://loadedtokens.com/companies/cartesia/): Speech models and voice agents, sold as monthly credits.
- [Cloudflare](https://loadedtokens.com/companies/cloudflare/): Workers AI inference and AI Gateway for managing model traffic.
- [Cognition](https://loadedtokens.com/companies/cognition/): Maker of Devin, the desktop and cloud coding agent sold on windsurf.com.
- [Comfy Org](https://loadedtokens.com/companies/comfy-org/): Makers of ComfyUI, a node-based local image and video workflow tool.
- [crewAI](https://loadedtokens.com/companies/crewai/): A framework for role-shaped agents written in Python.
- [Cursor](https://loadedtokens.com/companies/cursor/): An AI coding environment and command-line agent.
- [Deepgram](https://loadedtokens.com/companies/deepgram/): Speech-to-text, text-to-speech, and voice agents, billed on use.
- [DeepSeek](https://loadedtokens.com/companies/deepseek/): Open-weight language models and a hosted chat product.
- [ElevenLabs](https://loadedtokens.com/companies/elevenlabs/): Speech generation and Scribe speech-to-text tools.
- [Exa](https://loadedtokens.com/companies/exa/): A search API for agents, priced per request.
- [Flowise](https://loadedtokens.com/companies/flowise/): A visual builder for LLM flows you can host yourself.
- [GitHub](https://loadedtokens.com/companies/github/): Copilot coding assistance in editors, the cloud and the terminal.
- [Google](https://loadedtokens.com/companies/google/): Gemini developer models and AI Studio for prototyping with them.
- [HeyGen](https://loadedtokens.com/companies/heygen/): Avatar and translation video, with a credit system.
- [Hugging Face](https://loadedtokens.com/companies/hugging-face/): A model and dataset Hub with hosted inference provider access.
- [Jan](https://loadedtokens.com/companies/jan/): A desktop app for running local models, with an OpenAI-compatible local server.
- [LangChain](https://loadedtokens.com/companies/langchain/): LangGraph agent orchestration and LangSmith tracing and evaluation.
- [llama.cpp](https://loadedtokens.com/companies/llamacpp/): A local inference engine for running language models on CPUs and GPUs.
- [LM Studio](https://loadedtokens.com/companies/lmstudio/): A desktop app for running local models, with a chat window.
- [Luma AI](https://loadedtokens.com/companies/luma/): Hosted image and video generation, including Ray models.
- [Make](https://loadedtokens.com/companies/make/): A hosted automation tool that bills in credits per action.
- [Meta](https://loadedtokens.com/companies/meta/): Publisher of the Llama open-weight models.
- [Microsoft](https://loadedtokens.com/companies/microsoft/): Microsoft 365 Copilot for work and Copilot Studio for building agents.
- [MiniMax](https://loadedtokens.com/companies/minimax/): Publisher of MiniMax open-weight models and a hosted product.
- [Mistral AI](https://loadedtokens.com/companies/mistral/): Vibe assistant and APIs including document OCR.
- [Moonshot AI](https://loadedtokens.com/companies/moonshot/): Maker of Kimi, including open-weight Kimi checkpoints.
- [n8n](https://loadedtokens.com/companies/n8n/): A source-available workflow tool with AI nodes.
- [Ollama](https://loadedtokens.com/companies/ollama/): A local app and server for running open-weight models.
- [Open WebUI](https://loadedtokens.com/companies/open-webui/): A self-hosted chat interface for local and remote models.
- [OpenAI](https://loadedtokens.com/companies/openai/): ChatGPT and model APIs for writing, analysis and application development.
- [Perplexity](https://loadedtokens.com/companies/perplexity/): Model APIs with search tools, plus a standalone web-search API.
- [Replit](https://loadedtokens.com/companies/replit/): A hosted workspace that builds and runs apps from a prompt.
- [Runway](https://loadedtokens.com/companies/runway/): A hosted studio for generated video, images, and audio.
- [Stability AI](https://loadedtokens.com/companies/stability/): Publisher of Stable Diffusion and related image models.
- [Synthesia](https://loadedtokens.com/companies/synthesia/): Scripted avatar video for explainers and training.
- [Tavily](https://loadedtokens.com/companies/tavily/): A search API aimed at agents, with a free monthly credit.
- [Vercel](https://loadedtokens.com/companies/vercel/): The AI SDK for developers and v0 for generating and editing applications.
- [vLLM](https://loadedtokens.com/companies/vllm/): A server for serving open-weight models behind an API.
- [Zapier](https://loadedtokens.com/companies/zapier/): Workflow automation and MCP access to connected application actions.

## Published issues
- [The great localhost escape](https://loadedtokens.com/issues/2026-11-07-the-localhost-escape/): Cursor wants your side project to finally leave your laptop, Perplexity hands out agents with memory, and AWS sells answers by the call.
- [Four frontier models walk into a week](https://loadedtokens.com/issues/2026-10-30-model-week-from-heck/): OpenAI made smart cheap, Google made a model you can't have, DeepMind made a world, and Mistral just handed out the weights.

## Tool verdicts
- [Amazon Bedrock](https://loadedtokens.com/tools/amazon-bedrock/): Watch. AWS access to foundation models and supporting services such as knowledge bases and guardrails.
- [Amazon Q Developer](https://loadedtokens.com/tools/amazon-q-developer/): Use. A development assistant available in IDE and CLI workflows.
- [AssemblyAI](https://loadedtokens.com/tools/assemblyai/): Use. Speech-to-text with a published hourly rate, plus text to speech.
- [Cartesia](https://loadedtokens.com/tools/cartesia/): Watch. Speech models and voice agents sold as credits.
- [ChatGPT](https://loadedtokens.com/tools/chatgpt/): Use. A conversational assistant with file, research and writing features that depend on plan.
- [Claude](https://loadedtokens.com/tools/claude/): Use. A conversational assistant available on web, desktop and mobile with plan-dependent tools.
- [Claude Code](https://loadedtokens.com/tools/claude-code/): Watch. An agentic coding tool for repository and terminal workflows, described in its official overview.
- [Cloudflare AI Gateway](https://loadedtokens.com/tools/cloudflare-ai-gateway/): Use. Gateway controls including analytics, caching and rate limiting.
- [ComfyUI](https://loadedtokens.com/tools/comfyui/): Watch. A node graph for local image and video models.
- [Microsoft Copilot Studio](https://loadedtokens.com/tools/copilot-studio/): Watch. A platform for creating, deploying and managing custom agents connected to business workflows.
- [CrewAI](https://loadedtokens.com/tools/crewai/): Use. A Python framework for role-shaped agents.
- [Cursor CLI](https://loadedtokens.com/tools/cursor-cli/): Watch. A command-line agent with interactive and non-interactive workflows documented by Cursor.
- [Cursor editor](https://loadedtokens.com/tools/cursor-editor/): Watch. An AI coding environment with agent workflows and model selection.
- [Deepgram](https://loadedtokens.com/tools/deepgram/): Use. Speech-to-text, text-to-speech, and voice agents.
- [DeepSeek weights](https://loadedtokens.com/tools/deepseek-v4/): Watch. Open-weight DeepSeek models. The license lives on the model card.
- [Devin](https://loadedtokens.com/tools/devin/): Watch. A coding agent with a desktop app and cloud sessions.
- [ElevenLabs Scribe](https://loadedtokens.com/tools/elevenlabs-scribe/): Watch. Speech-to-text tools for recorded and real-time transcription, with timestamps and speaker features.
- [ElevenLabs Text to Speech](https://loadedtokens.com/tools/elevenlabs-tts/): Watch. Speech generation with plan-dependent voice features, credit usage and licensing.
- [Exa](https://loadedtokens.com/tools/exa/): Use. A search API for agents, with contents and an answer endpoint.
- [Flowise](https://loadedtokens.com/tools/flowise/): Use. A visual builder for LLM flows you host yourself.
- [FLUX.1 schnell](https://loadedtokens.com/tools/flux-schnell/): Use. The Apache-licensed FLUX.1 schnell image model.
- [Gemini Developer API](https://loadedtokens.com/tools/gemini-api/): Watch. A developer API with model-specific free and paid access.
- [Gemini Omni](https://loadedtokens.com/tools/gemini-omni/): Use. Google's native video model in the Gemini API.
- [GitHub Copilot](https://loadedtokens.com/tools/github-copilot/): Use. Coding assistance with completions, agent features and model access that vary by plan.
- [GitHub Copilot CLI](https://loadedtokens.com/tools/github-copilot-cli/): Watch. A terminal-based coding agent documented by GitHub.
- [Google AI Studio](https://loadedtokens.com/tools/google-ai-studio/): Use. Google’s browser-based starting point for exploring Gemini and developing API-backed workflows.
- [GPT-Live](https://loadedtokens.com/tools/gpt-live/): Watch. OpenAI's live voice session, including outbound SIP.
- [HeyGen](https://loadedtokens.com/tools/heygen/): Watch. Avatar video and translation.
- [Hugging Face Inference Providers](https://loadedtokens.com/tools/hf-inference-providers/): Watch. Hosted access to inference providers with centralized pay-as-you-go billing.
- [Hugging Face Hub](https://loadedtokens.com/tools/hugging-face-hub/): Use. A Hub for discovering and sharing models, datasets and Spaces.
- [Jan](https://loadedtokens.com/tools/jan/): Watch. A desktop app that runs local models and exposes a local API.
- [Kimi](https://loadedtokens.com/tools/kimi/): Watch. Moonshot's Kimi models, including open-weight checkpoints.
- [LangGraph](https://loadedtokens.com/tools/langgraph/): Use. An orchestration framework for stateful agent workflows, with explicit graph structure and execution control.
- [LangSmith](https://loadedtokens.com/tools/langsmith/): Use. A platform for tracing, evaluation and agent-development workflows.
- [Llama](https://loadedtokens.com/tools/llama/): Watch. Meta's Llama open-weight models.
- [llama.cpp](https://loadedtokens.com/tools/llama-cpp/): Use. A local inference engine for GGUF-style model files.
- [LM Studio](https://loadedtokens.com/tools/lm-studio/): Watch. A desktop app for chatting with local models.
- [Luma Dream Machine](https://loadedtokens.com/tools/luma-dream-machine/): Watch. Hosted video and image generation.
- [Make](https://loadedtokens.com/tools/make/): Watch. Hosted scenarios that connect apps, billed in credits.
- [Microsoft 365 Copilot](https://loadedtokens.com/tools/microsoft-365-copilot/): Watch. Workplace AI integrated with Microsoft 365 applications, with license-dependent capabilities.
- [MiniMax weights](https://loadedtokens.com/tools/minimax-weights/): Watch. MiniMax open-weight checkpoints.
- [Mistral OCR](https://loadedtokens.com/tools/mistral-ocr/): Use. A document OCR API for extracting content from documents, documented in Mistral’s document AI guide.
- [Mistral open weights](https://loadedtokens.com/tools/mistral-open/): Watch. Mistral models you can download, as distinct from the paid API.
- [Mistral Studio](https://loadedtokens.com/tools/mistral-studio/): Watch. Mistral's hosted place to run agents and workflows.
- [Mistral Vibe](https://loadedtokens.com/tools/mistral-vibe/): Watch. An assistant experience with chat, search, creation and coding access described on the pricing page.
- [n8n](https://loadedtokens.com/tools/n8n/): Watch. A workflow tool with AI nodes. Source-available, not OSI open source.
- [Ollama](https://loadedtokens.com/tools/ollama/): Use. A local runner for open-weight chat models.
- [Open WebUI](https://loadedtokens.com/tools/open-webui/): Watch. A chat interface you host in front of Ollama or another model server.
- [OpenAI Responses API](https://loadedtokens.com/tools/openai-responses-api/): Use. A model API surface with usage billed at the selected model and tool rates.
- [Perplexity Agent API](https://loadedtokens.com/tools/perplexity-agent-api/): Watch. An API providing third-party model access with token-based billing and separately priced tools.
- [Perplexity Search API](https://loadedtokens.com/tools/perplexity-search-api/): Use. A web-search API with separate request-based billing listed in Perplexity’s rate card.
- [Perplexity Teammates](https://loadedtokens.com/tools/perplexity-teammates/): Watch. Personal agents that run long projects, remember goals and files across runs, and report back.
- [Qwen](https://loadedtokens.com/tools/qwen/): Watch. Alibaba's Qwen open-weight family.
- [Replit Agent](https://loadedtokens.com/tools/replit-agent/): Watch. A hosted agent that builds and deploys apps inside Replit.
- [Runway](https://loadedtokens.com/tools/runway/): Watch. A hosted studio for AI video and images.
- [Stable Diffusion](https://loadedtokens.com/tools/stable-diffusion/): Watch. The Stable Diffusion image models.
- [Synthesia](https://loadedtokens.com/tools/synthesia/): Watch. Scripted avatar video.
- [Tavily](https://loadedtokens.com/tools/tavily/): Watch. Search API credits for agents.
- [v0](https://loadedtokens.com/tools/v0/): Watch. A generation and editing product with visual design, code and Vercel deployment workflows.
- [Vercel AI SDK](https://loadedtokens.com/tools/vercel-ai-sdk/): Use. A TypeScript toolkit with provider abstractions and UI building blocks for AI applications.
- [vLLM](https://loadedtokens.com/tools/vllm/): Use. An Apache-licensed server for open-weight models.
- [Whisper](https://loadedtokens.com/tools/whisper/): Use. OpenAI's open speech-recognition models and code.
- [Cloudflare Workers AI](https://loadedtokens.com/tools/workers-ai/): Use. Managed model inference integrated with the Workers platform.
- [Zapier MCP](https://loadedtokens.com/tools/zapier-mcp/): Watch. An MCP connection layer for actions in applications connected to Zapier.
- [Zapier workflows](https://loadedtokens.com/tools/zapier-workflows/): Use. Task-based automation combining app workflows, data tables and forms.
- [Claude Code 2.1.319](https://loadedtokens.com/tools/claude-code-2-1-319/): Use (evidence limited). Background MCP tasks and a permissions.denyInsteadOfPrompting setting for unattended runs.
- [Web Answer on Bedrock AgentCore](https://loadedtokens.com/tools/aws-web-answer/): Watch. One call returns a synthesized, cited web answer for your agent, over MCP.
- [Codex CLI 0.167](https://loadedtokens.com/tools/codex-cli-0-167/): Use (evidence limited). An agent command center (codex agents) that flags unread results across several agents.
- [Cursor Live Apps](https://loadedtokens.com/tools/cursor-live-apps/): Watch. A cloud agent builds the app, sets up a database and auth, and publishes it to a shareable link.
- [Tencent Hy4](https://loadedtokens.com/tools/tencent-hy4/): Watch. Open-source-bound model with vision and computer use; weights promised "in the coming days."
- [Gemini 4 Carbon](https://loadedtokens.com/tools/gemini-4-carbon/): Watch. Google's strongest model for cybersecurity and long-horizon tasks.
- [Genie 4](https://loadedtokens.com/tools/genie-4/): Ignore. Real-time interactive world simulation with native audio.
- [GPT-6.1 Terra](https://loadedtokens.com/tools/gpt-6-1-terra/): Use (evidence limited). Fast, cheaper model for computer use, visual reasoning, writing and coding.

## Methodology
- [About and corrections](https://loadedtokens.com/about/)