--- title: LocateAnything-3B, Qwen-AgentWorld-35B-A3B, and... description: Three new AI foundation models are now available on Amazon SageMaker JumpStart: NVIDIA's LocateAnything-3B for visual grounding and object localization using a... canonical: https://daily.dev/posts/locateanything-3b-qwen-agentworld-35b-a3b-and-qwen3-5-122b-a10b-models-now-available-on-amazon-sag-wzuczap31 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: LocateAnything-3B, Qwen-AgentWorld-35B-A3B, and Qwen3.5-122B-A10B models now available on Amazon SageMaker JumpStart | daily.dev og:description: Three new AI foundation models are now available on Amazon SageMaker JumpStart: NVIDIA's LocateAnything-3B for visual grounding and object localization using a... og:url: https://daily.dev/posts/locateanything-3b-qwen-agentworld-35b-a3b-and-qwen3-5-122b-a10b-models-now-available-on-amazon-sag-wzuczap31 og:image: https://api.daily.dev/og/posts/wZUczaP31.png og:image:alt: LocateAnything-3B, Qwen-AgentWorld-35B-A3B, and Qwen3.5-122B-A10B models now available on Amazon SageMaker JumpStart og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # LocateAnything-3B, Qwen-AgentWorld-35B-A3B, and Qwen3.5-122B-A10B models now available on Amazon SageMaker JumpStart **[AWS](https://daily.dev/sources/aws)** · 2 min read · 2 upvotes · 0 comments ## Summary Three new AI foundation models are now available on Amazon SageMaker JumpStart: NVIDIA's LocateAnything-3B for visual grounding and object localization using a Parallel Box Decoding framework; Qwen-AgentWorld-35B-A3B, the first language world model covering seven agent interaction domains (tool calling, search, terminal, software engineering, Android, web, and OS), trained on over 10 million real-world trajectories; and Qwen3.5-122B-A10B, a 122B-parameter multimodal model with only 10B activated per token via a hybrid Gated Delta Networks and sparse MoE architecture, featuring a 262K context window. All three can be deployed via the SageMaker JumpStart console or Python SDK. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is Qwen3.5-122B-A10B and how does its architecture differ from a standard dense model? Qwen3.5-122B-A10B is a multimodal reasoning model with 122B total parameters but only 10B activated per token, using a hybrid architecture that combines Gated Delta Networks with a sparse Mixture-of-Experts layer of 256 experts. This design delivers strong reasoning, coding, agent, and visual understanding performance with a native 262K token context window and minimal latency overhead compared to a fully dense model of similar size. _Teams evaluating efficient large-scale reasoning models can track releases like this one on daily.dev._ ### What does Qwen-AgentWorld-35B-A3B simulate and what domains does it cover? Qwen-AgentWorld-35B-A3B is a language world model that predicts next environment states given an agent's action and interaction history, using long chain-of-thought reasoning. It is the first such model to cover all seven interaction domains in one model: tool calling, search, terminal, software engineering, Android, web, and OS interaction, trained on over 10 million real-world interaction trajectories. _Developers building agent simulation pipelines can follow model updates like this via daily.dev._ ### How does LocateAnything-3B perform visual grounding differently from typical object detection models? LocateAnything-3B uses a Parallel Box Decoding (PBD) framework that decodes bounding boxes and points as atomic units in a single step rather than sequentially, preserving geometric coherence while enabling substantial parallelism. This allows precise object localization, dense detection, and point-based localization from natural language instructions across enterprise and physical AI applications. _Engineers comparing visual grounding approaches can keep tabs on new releases like this through daily.dev._ ## Similar posts on daily.dev - [NVIDIA Nemotron-3-Super-120B, Qwen3.5-9B, and Qwen3.5-27B models now available on Amazon SageMaker JumpStart](https://daily.dev/posts/nvidia-nemotron-3-super-120b-qwen3-5-9b-and-qwen3-5-27b-models-now-available-on-amazon-sagemaker-j-grcvqwjnf) · AWS · 1 upvotes · 0 comments - [DeepSeek OCR, MiniMax M2.1, and Qwen3-VL-8B-Instruct models are now available on SageMaker JumpStart](https://daily.dev/posts/deepseek-ocr-minimax-m2-1-and-qwen3-vl-8b-instruct-models-are-now-available-on-sagemaker-jumpstart-weneybp9z) · AWS · 1 upvotes · 0 comments - [Qwen3.6–35B-A3B: The Most Practical Open-Source AI Model Yet?](https://daily.dev/posts/qwen3-6-35b-a3b-the-most-practical-open-source-ai-model-yet--eihfhk9n5) · Faun · 77 upvotes · 6 comments - [Cosmos3-Edge, Cosmos3-Nano, and Cosmos3-Super models now available on Amazon SageMaker JumpStart](https://daily.dev/posts/cosmos3-edge-cosmos3-nano-and-cosmos3-super-models-now-available-on-amazon-sagemaker-jumpstart-4ec9r48zh) · AWS · 1 upvotes · 1 comments --- Tags: [#ai](https://daily.dev/tags/ai), [#aws](https://daily.dev/tags/aws), [#multimodal](https://daily.dev/tags/multimodal), [#mixture-of-experts](https://daily.dev/tags/mixture-of-experts) [View this post on daily.dev](https://daily.dev/posts/locateanything-3b-qwen-agentworld-35b-a3b-and-qwen3-5-122b-a10b-models-now-available-on-amazon-sag-wzuczap31) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"LocateAnything-3B, Qwen-AgentWorld-35B-A3B, and Qwen3.5-122B-A10B models now available on Amazon SageMaker JumpStart","url":"https://daily.dev/posts/locateanything-3b-qwen-agentworld-35b-a3b-and-qwen3-5-122b-a10b-models-now-available-on-amazon-sag-wzuczap31","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/locateanything-3b-qwen-agentworld-35b-a3b-and-qwen3-5-122b-a10b-models-now-available-on-amazon-sag-wzuczap31"},"datePublished":"2026-08-11T16:03:10.018Z","dateModified":"2026-09-13T18:55:08.588Z","description":"Three new AI foundation models are now available on Amazon SageMaker JumpStart: NVIDIA's LocateAnything-3B for visual grounding and object localization using a...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/2d99c2aecc14a73a5cba88addc902c88?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/2d99c2aecc14a73a5cba88addc902c88?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"AWS","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"AWS","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/aws","url":"https://daily.dev/sources/aws"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/locateanything-3b-qwen-agentworld-35b-a3b-and-qwen3-5-122b-a10b-models-now-available-on-amazon-sag-wzuczap31","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai,aws,multimodal,mixture-of-experts","timeRequired":"PT2M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"AWS","item":"https://daily.dev/sources/aws"},{"@type":"ListItem","position":3,"name":"LocateAnything-3B, Qwen-AgentWorld-35B-A3B, and Qwen3.5-122B-A10B models now available on Amazon SageMaker JumpStart"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/locateanything-3b-qwen-agentworld-35b-a3b-and-qwen3-5-122b-a10b-models-now-available-on-amazon-sag-wzuczap31#faq","mainEntity":[{"@type":"Question","name":"What is Qwen3.5-122B-A10B and how does its architecture differ from a standard dense model?","acceptedAnswer":{"@type":"Answer","text":"Qwen3.5-122B-A10B is a multimodal reasoning model with 122B total parameters but only 10B activated per token, using a hybrid architecture that combines Gated Delta Networks with a sparse Mixture-of-Experts layer of 256 experts. This design delivers strong reasoning, coding, agent, and visual understanding performance with a native 262K token context window and minimal latency overhead compared to a fully dense model of similar size. Teams evaluating efficient large-scale reasoning models can track releases like this one on daily.dev."}},{"@type":"Question","name":"What does Qwen-AgentWorld-35B-A3B simulate and what domains does it cover?","acceptedAnswer":{"@type":"Answer","text":"Qwen-AgentWorld-35B-A3B is a language world model that predicts next environment states given an agent's action and interaction history, using long chain-of-thought reasoning. It is the first such model to cover all seven interaction domains in one model: tool calling, search, terminal, software engineering, Android, web, and OS interaction, trained on over 10 million real-world interaction trajectories. Developers building agent simulation pipelines can follow model updates like this via daily.dev."}},{"@type":"Question","name":"How does LocateAnything-3B perform visual grounding differently from typical object detection models?","acceptedAnswer":{"@type":"Answer","text":"LocateAnything-3B uses a Parallel Box Decoding (PBD) framework that decodes bounding boxes and points as atomic units in a single step rather than sequentially, preserving geometric coherence while enabling substantial parallelism. This allows precise object localization, dense detection, and point-based localization from natural language instructions across enterprise and physical AI applications. Engineers comparing visual grounding approaches can keep tabs on new releases like this through daily.dev."}}]} ```