{"id":44,"date":"2026-09-02T19:58:45","date_gmt":"2026-09-02T11:58:45","guid":{"rendered":"http:\/\/localhost:8888\/?page_id=44"},"modified":"2026-09-04T20:19:01","modified_gmt":"2026-09-04T12:19:01","slug":"research","status":"publish","type":"page","link":"https:\/\/tongyi.ai\/?page_id=44","title":{"rendered":"RESEARCH"},"content":{"rendered":"\n<div class=\"wp-block-columns alignwide are-vertically-aligned-center is-not-stacked-on-mobile is-layout-flex wp-container-core-columns-is-layout-8f761849 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-vertically-aligned-center is-layout-flow wp-block-column-is-layout-flow\"><div class=\"wp-block-image\">\n<figure class=\"alignleft size-full is-resized\"><img loading=\"lazy\" decoding=\"async\" width=\"1869\" height=\"501\" src=\"http:\/\/30.48.154.101:8888\/wp-content\/uploads\/2026\/09\/\u8d44\u6e90-1@4x-1.png\" alt=\"\" class=\"wp-image-200\" style=\"aspect-ratio:3.730841121495327;width:195px;height:auto\" srcset=\"https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/\u8d44\u6e90-1@4x-1.png 1869w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/\u8d44\u6e90-1@4x-1-300x80.png 300w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/\u8d44\u6e90-1@4x-1-1024x274.png 1024w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/\u8d44\u6e90-1@4x-1-768x206.png 768w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/\u8d44\u6e90-1@4x-1-1536x412.png 1536w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/\u8d44\u6e90-1@4x-1-938x251.png 938w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/\u8d44\u6e90-1@4x-1-1130x303.png 1130w\" sizes=\"auto, (max-width: 1869px) 100vw, 1869px\" \/><\/figure>\n<\/div><\/div>\n\n\n\n<div class=\"wp-block-column is-vertically-aligned-center is-layout-flow wp-block-column-is-layout-flow\"><nav class=\"is-responsive items-justified-center alignwide wp-block-navigation is-horizontal is-content-justification-center is-layout-flex wp-container-core-navigation-is-layout-7d812b4c wp-block-navigation-is-layout-flex\" aria-label=\"\u5bfc\u822a\u680f\" \n\t\t data-wp-interactive=\"core\/navigation\" data-wp-context='{\"overlayOpenedBy\":{\"click\":false,\"hover\":false,\"focus\":false},\"type\":\"overlay\",\"roleAttribute\":\"\",\"ariaLabel\":\"Menu\"}'><button aria-haspopup=\"dialog\" aria-label=\"Open menu\" class=\"wp-block-navigation__responsive-container-open\" \n\t\t\t\tdata-wp-on--click=\"actions.openMenuOnClick\"\n\t\t\t\tdata-wp-on--keydown=\"actions.handleMenuKeydown\"\n\t\t\t><svg width=\"24\" height=\"24\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" viewBox=\"0 0 24 24\"><path d=\"M5 5v1.5h14V5H5z\"><\/path><path d=\"M5 12.8h14v-1.5H5v1.5z\"><\/path><path d=\"M5 19h14v-1.5H5V19z\"><\/path><\/svg><\/button>\n\t\t\t\t<div class=\"wp-block-navigation__responsive-container\"  id=\"modal-1\" \n\t\t\t\tdata-wp-class--has-modal-open=\"state.isMenuOpen\"\n\t\t\t\tdata-wp-class--is-menu-open=\"state.isMenuOpen\"\n\t\t\t\tdata-wp-watch=\"callbacks.initMenu\"\n\t\t\t\tdata-wp-on--keydown=\"actions.handleMenuKeydown\"\n\t\t\t\tdata-wp-on--focusout=\"actions.handleMenuFocusout\"\n\t\t\t\ttabindex=\"-1\"\n\t\t\t>\n\t\t\t\t\t<div class=\"wp-block-navigation__responsive-close\" tabindex=\"-1\">\n\t\t\t\t\t\t<div class=\"wp-block-navigation__responsive-dialog\" \n\t\t\t\tdata-wp-bind--aria-modal=\"state.ariaModal\"\n\t\t\t\tdata-wp-bind--aria-label=\"state.ariaLabel\"\n\t\t\t\tdata-wp-bind--role=\"state.roleAttribute\"\n\t\t\t>\n\t\t\t\t\t\t\t<button aria-label=\"Close menu\" class=\"wp-block-navigation__responsive-container-close\" \n\t\t\t\tdata-wp-on--click=\"actions.closeMenuOnClick\"\n\t\t\t><svg xmlns=\"http:\/\/www.w3.org\/2000\/svg\" viewBox=\"0 0 24 24\" width=\"24\" height=\"24\" aria-hidden=\"true\" focusable=\"false\"><path d=\"m13.06 12 6.47-6.47-1.06-1.06L12 10.94 5.53 4.47 4.47 5.53 10.94 12l-6.47 6.47 1.06 1.06L12 13.06l6.47 6.47 1.06-1.06L13.06 12Z\"><\/path><\/svg><\/button>\n\t\t\t\t\t\t\t<div class=\"wp-block-navigation__responsive-container-content\" \n\t\t\t\tdata-wp-watch=\"callbacks.focusFirstElement\"\n\t\t\t id=\"modal-1-content\">\n\t\t\t\t\t\t\t\t\n<div class=\"wp-block-buttons wp-container-content-9cfa9a5a is-horizontal is-content-justification-right is-layout-flex wp-container-core-buttons-is-layout-567e7cbc wp-block-buttons-is-layout-flex\">\n<div class=\"wp-block-button\"><a class=\"wp-block-button__link wp-element-button\" href=\"https:\/\/tongyi.ai\/\">Transparency<\/a><\/div>\n\n\n\n<div class=\"wp-block-button\"><a class=\"wp-block-button__link wp-element-button\" href=\"https:\/\/tongyi.ai\/?page_id=44\"> Research<\/a><\/div>\n<\/div>\n\n\t\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div><\/nav><\/div>\n<\/div>\n\n\n<h2 style=\"font-size:50px\" class=\"alignwide wp-block-post-title\">RESEARCH<\/h2>\n\n<ul class=\"wp-block-latest-posts__list is-grid columns-2 has-dates alignwide wp-block-latest-posts has-medium-font-size has-alibaba-puhuiti-2-0-font-family is-layout-grid wp-container-core-latest-posts-is-layout-28b052f3 wp-block-latest-posts-is-layout-grid\"><li><div class=\"wp-block-latest-posts__featured-image\"><a href=\"https:\/\/tongyi.ai\/?p=236\" aria-label=\"E-Commerce Bench: Long-Horizon Operations, Multi-Dimensional Evaluation\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"614\" src=\"https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01QlThshCJ7uE3E9bs_6000000000358-2-tps-1590-954-1024x614.webp\" class=\"attachment-large size-large wp-post-image\" alt=\"\" style=\"\" srcset=\"https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01QlThshCJ7uE3E9bs_6000000000358-2-tps-1590-954-1024x614.webp 1024w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01QlThshCJ7uE3E9bs_6000000000358-2-tps-1590-954-300x180.webp 300w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01QlThshCJ7uE3E9bs_6000000000358-2-tps-1590-954-768x461.webp 768w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01QlThshCJ7uE3E9bs_6000000000358-2-tps-1590-954-1536x922.webp 1536w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01QlThshCJ7uE3E9bs_6000000000358-2-tps-1590-954-938x563.webp 938w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01QlThshCJ7uE3E9bs_6000000000358-2-tps-1590-954-1130x678.webp 1130w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01QlThshCJ7uE3E9bs_6000000000358-2-tps-1590-954.webp 1590w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/a><\/div><a class=\"wp-block-latest-posts__post-title\" href=\"https:\/\/tongyi.ai\/?p=236\">E-Commerce Bench: Long-Horizon Operations, Multi-Dimensional Evaluation<\/a><time datetime=\"2026-09-04T11:20:16+08:00\" class=\"wp-block-latest-posts__post-date\">September 4, 2026<\/time><div class=\"wp-block-latest-posts__post-excerpt\">Agent benchmarks over the past few years have mostly followed one pattern. A goal is handed to the model, and the model tries to reach it within a bounded number of turns, whether that means finding the treasure in a maze, producing a report, or fixing a piece of code. Performance is then scored on the quality of the deliverable or on how much of the task got done, and evaluations of this kind usually come with a well-defined natural stopping point. Most long-horizon tasks\u2026 <a class=\"wp-block-latest-posts__read-more\" href=\"https:\/\/tongyi.ai\/?p=236\" rel=\"noopener\">Read more<span class=\"screen-reader-text\">: E-Commerce Bench: Long-Horizon Operations, Multi-Dimensional Evaluation<\/span><\/a><\/div><\/li>\n<li><div class=\"wp-block-latest-posts__featured-image\"><a href=\"https:\/\/tongyi.ai\/?p=233\" aria-label=\"Qwen-Drive-1.0: An Initial Step towards a Vision-Language Foundation Model for Autonomous Driving\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"614\" src=\"https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01BEtEP74nVrF3E9bs_6000000002925-2-tps-1590-954-1024x614.jpg\" class=\"attachment-large size-large wp-post-image\" alt=\"\" style=\"\" srcset=\"https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01BEtEP74nVrF3E9bs_6000000002925-2-tps-1590-954-1024x614.jpg 1024w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01BEtEP74nVrF3E9bs_6000000002925-2-tps-1590-954-300x180.jpg 300w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01BEtEP74nVrF3E9bs_6000000002925-2-tps-1590-954-768x461.jpg 768w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01BEtEP74nVrF3E9bs_6000000002925-2-tps-1590-954-1536x922.jpg 1536w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01BEtEP74nVrF3E9bs_6000000002925-2-tps-1590-954-938x563.jpg 938w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01BEtEP74nVrF3E9bs_6000000002925-2-tps-1590-954-1130x678.jpg 1130w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01BEtEP74nVrF3E9bs_6000000002925-2-tps-1590-954.jpg 1590w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/a><\/div><a class=\"wp-block-latest-posts__post-title\" href=\"https:\/\/tongyi.ai\/?p=233\">Qwen-Drive-1.0: An Initial Step towards a Vision-Language Foundation Model for Autonomous Driving<\/a><time datetime=\"2026-09-04T11:18:49+08:00\" class=\"wp-block-latest-posts__post-date\">September 4, 2026<\/time><div class=\"wp-block-latest-posts__post-excerpt\">Introduction We introduce Qwen-Drive-1.0,&nbsp;the first vision-language foundation model for autonomous driving that unifies 3D perception and visual question answering at the pretraining stage and further extends to motion planning,&nbsp;while keeping the pretrained VLM architecture entirely untouched. Built on the natively multimodal Qwen3.5-4B, it attaches two external modules. A BEV perception head serves as an explicit, inspectable 3D probe, jointly performing 3D object detection, semantic occupancy prediction, and BEV map segmentation, and a Planning Expert generates future ego trajectories through flow matching. Through staged training, we\u2026 <a class=\"wp-block-latest-posts__read-more\" href=\"https:\/\/tongyi.ai\/?p=233\" rel=\"noopener\">Read more<span class=\"screen-reader-text\">: Qwen-Drive-1.0: An Initial Step towards a Vision-Language Foundation Model for Autonomous Driving<\/span><\/a><\/div><\/li>\n<li><div class=\"wp-block-latest-posts__featured-image\"><a href=\"https:\/\/tongyi.ai\/?p=226\" aria-label=\"Qwen3.8-Max: A New Bar for Coding and Cowork\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"614\" src=\"https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ITn7j2cu9OF3E9bs_6000000000314-2-tps-1590-954-1-1024x614.jpg\" class=\"attachment-large size-large wp-post-image\" alt=\"\" style=\"\" srcset=\"https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ITn7j2cu9OF3E9bs_6000000000314-2-tps-1590-954-1-1024x614.jpg 1024w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ITn7j2cu9OF3E9bs_6000000000314-2-tps-1590-954-1-300x180.jpg 300w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ITn7j2cu9OF3E9bs_6000000000314-2-tps-1590-954-1-768x461.jpg 768w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ITn7j2cu9OF3E9bs_6000000000314-2-tps-1590-954-1-1536x922.jpg 1536w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ITn7j2cu9OF3E9bs_6000000000314-2-tps-1590-954-1-938x563.jpg 938w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ITn7j2cu9OF3E9bs_6000000000314-2-tps-1590-954-1-1130x678.jpg 1130w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ITn7j2cu9OF3E9bs_6000000000314-2-tps-1590-954-1.jpg 1590w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/a><\/div><a class=\"wp-block-latest-posts__post-title\" href=\"https:\/\/tongyi.ai\/?p=226\">Qwen3.8-Max: A New Bar for Coding and Cowork<\/a><time datetime=\"2026-09-04T11:14:02+08:00\" class=\"wp-block-latest-posts__post-date\">September 4, 2026<\/time><div class=\"wp-block-latest-posts__post-excerpt\">Today, we are officially releasing&nbsp;Qwen 3.8-Max, the most capable model in the Qwen family to date. This also marks the first time we will open-source the weights of a Qwen-Max-class model \u2014 the open weights will be released next week. Built upon the architectural foundation of Qwen 3.5, Qwen 3.8-Max scales to&nbsp;2.4 trillion&nbsp;parameters, delivering comprehensive improvements across coding, work, research, and long-horizon tasks. It can not only answer more challenging questions, but also complete complex tasks end-to-end with greater reliability, producing dependable deliverables. Coding For\u2026 <a class=\"wp-block-latest-posts__read-more\" href=\"https:\/\/tongyi.ai\/?p=226\" rel=\"noopener\">Read more<span class=\"screen-reader-text\">: Qwen3.8-Max: A New Bar for Coding and Cowork<\/span><\/a><\/div><\/li>\n<li><div class=\"wp-block-latest-posts__featured-image\"><a href=\"https:\/\/tongyi.ai\/?p=227\" aria-label=\"Qwen3.8-Flash-Next: A New Architecture, Towards Ultimate Cost-Efficiency\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"614\" src=\"https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ZxcR5nZsP9E3E9bs_6000000003536-2-tps-1590-954-1-1024x614.jpg\" class=\"attachment-large size-large wp-post-image\" alt=\"\" style=\"\" srcset=\"https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ZxcR5nZsP9E3E9bs_6000000003536-2-tps-1590-954-1-1024x614.jpg 1024w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ZxcR5nZsP9E3E9bs_6000000003536-2-tps-1590-954-1-300x180.jpg 300w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ZxcR5nZsP9E3E9bs_6000000003536-2-tps-1590-954-1-768x461.jpg 768w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ZxcR5nZsP9E3E9bs_6000000003536-2-tps-1590-954-1-1536x922.jpg 1536w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ZxcR5nZsP9E3E9bs_6000000003536-2-tps-1590-954-1-938x563.jpg 938w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ZxcR5nZsP9E3E9bs_6000000003536-2-tps-1590-954-1-1130x678.jpg 1130w, https:\/\/tongyi.ai\/wp-content\/uploads\/2026\/09\/O1CN01ZxcR5nZsP9E3E9bs_6000000003536-2-tps-1590-954-1.jpg 1590w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/a><\/div><a class=\"wp-block-latest-posts__post-title\" href=\"https:\/\/tongyi.ai\/?p=227\">Qwen3.8-Flash-Next: A New Architecture, Towards Ultimate Cost-Efficiency<\/a><time datetime=\"2026-09-04T11:12:58+08:00\" class=\"wp-block-latest-posts__post-date\">September 4, 2026<\/time><div class=\"wp-block-latest-posts__post-excerpt\">Introduction In this release we are opening the weights of&nbsp;Qwen3.8-Flash-Next, a multimodal MoE model that also serves as an early preview of the architecture used in&nbsp;Qwen4. It plays the same role that&nbsp;Qwen3-Next&nbsp;played for Qwen3.5: the hybrid&nbsp;Gated DeltaNet + Gated Attention&nbsp;design introduced at that time has since been used across the Qwen3.5, Qwen3.6, Qwen3.7 and Qwen3.8 series. We are again releasing the architectural changes early, so that the community can examine them before the full Qwen4 model family is built on top of them. Qwen3.8-Flash-Next upgrades\u2026 <a class=\"wp-block-latest-posts__read-more\" href=\"https:\/\/tongyi.ai\/?p=227\" rel=\"noopener\">Read more<span class=\"screen-reader-text\">: Qwen3.8-Flash-Next: A New Architecture, Towards Ultimate Cost-Efficiency<\/span><\/a><\/div><\/li>\n<\/ul>","protected":false},"excerpt":{"rendered":"","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":1,"comment_status":"closed","ping_status":"closed","template":"template-front-page.php","meta":{"footnotes":""},"class_list":["post-44","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/tongyi.ai\/index.php?rest_route=\/wp\/v2\/pages\/44","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/tongyi.ai\/index.php?rest_route=\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/tongyi.ai\/index.php?rest_route=\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/tongyi.ai\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/tongyi.ai\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=44"}],"version-history":[{"count":28,"href":"https:\/\/tongyi.ai\/index.php?rest_route=\/wp\/v2\/pages\/44\/revisions"}],"predecessor-version":[{"id":279,"href":"https:\/\/tongyi.ai\/index.php?rest_route=\/wp\/v2\/pages\/44\/revisions\/279"}],"wp:attachment":[{"href":"https:\/\/tongyi.ai\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=44"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}