--- format: "aidr-story-markdown/v1" id: "ecb79e81631a0d6bc8cbdb2375fa0230523b3ad3716482c0cf05f2e192ac5601" canonical_url: "https://aidr.today/ecb79e81?lang=en" title: "Tencent Releases Hy-Nomas Video World Model and Interaction Benchmark" lang: "en" requested_lang: "en" available_langs: ["en","vi"] translation_fallback: null fallback_fields: [] published_at: "2026-10-23T15:38:56.000Z" category: "Open Source" topics: ["open-source","llm"] source_urls: ["https://marketbrief.now/ai/tencent-releases-hy-nomas-video-world-model-and-interaction-benchmark-65fbb5dd","https://huggingnews.com/ai/tencent-releases-hy-nomas-video-world-model-and-interaction-benchmark-65fbb5dd"] summary: "Tencent's Hunyuan division released the Hy-Nomas video world model, which supports multi-turn text and camera control while producing 720p video at 24 FPS. The model is available under an Apache 2.0 license and uses KV cache memory for continuous generation, with an inference option compatible with a single RTX 4090. Tencent also published Hy-Nomas-Bench on Hugging Face, a benchmark with 280 test cases for evaluating controllable video world models." --- # Tencent Releases Hy\-Nomas Video World Model and Interaction Benchmark > [Open the canonical story]() **Published:** 2026-10-23T15:38:56.000Z **Category:** Open Source **Topics:** open\-source, llm ## Summary Tencent's Hunyuan division released the Hy\-Nomas video world model, which supports multi\-turn text and camera control while producing 720p video at 24 FPS\. The model is available under an Apache 2\.0 license and uses KV cache memory for continuous generation, with an inference option compatible with a single RTX 4090\. Tencent also published Hy\-Nomas\-Bench on Hugging Face, a benchmark with 280 test cases for evaluating controllable video world models\. ## Sources - [Story source]() - [Story source]()