--- title: Qwen 3.8 27B: 16GB VRAM Local Test | daily.dev description: A hands-on test of the newly released Qwen 3.8 27B model running locally on a 16GB VRAM RTX 5060Ti via llama.cpp. The dense hybrid model (linear attention on... canonical: https://daily.dev/posts/zhgzjwota twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Qwen 3.8 27B: 16GB VRAM Local Test | daily.dev og:description: A hands-on test of the newly released Qwen 3.8 27B model running locally on a 16GB VRAM RTX 5060Ti via llama.cpp. The dense hybrid model (linear attention on... og:url: https://daily.dev/posts/zhgzjwota og:image: https://api.daily.dev/og/posts/ZhgzJwoTA.png og:image:alt: Post cover image 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. # Qwen 3.8 27B: 16GB VRAM Local Test **[nerdalytics](https://daily.dev/sources/jcwojcedqzpspqbvk73nc)** · [@nerdalytics](https://daily.dev/nerdalytics) · 25 upvotes · 6 comments ## Summary A hands-on test of the newly released Qwen 3.8 27B model running locally on a 16GB VRAM RTX 5060Ti via llama.cpp. The dense hybrid model (linear attention on 48 of 64 layers plus gated attention) supports native multimodal input including video, a 262K context window expandable to 1M tokens, and toggleable thinking mode. Coding tasks like a portfolio site and a 3D racing game were generated but were slow (4.78-6.42 tokens/sec) and required long generation times (over an hour). Benchmark scores are compared against Opus 4.6 Max, showing gains on terminal-bench, LiveCodeBench, DeepSeek 1.1, and OSWorld Verified. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Community discussion Top comments from developers on daily.dev. **@lezli01** · 4 upvotes > I really do not see LLMs of these size to be reasonable for agentic workload. **@agustinbarrientos** · 0 upvotes > Could you rerun the portfolio prompt with a quant that fits entirely in 16 GB of VRAM? **@lucasgoularte** · 0 upvotes > I really loved this thumbnail ## Similar posts on daily.dev - [Qwen 3.6 27B is the sweet spot for local development](https://daily.dev/posts/qwen-3-6-27b-is-the-sweet-spot-for-local-development-iqrlzajwn) · Quesma · 2 upvotes · 0 comments - [I tested 3 local LLMs on my RTX 4070 Ti for real work — only one earned a permanent spot](https://daily.dev/posts/i-tested-3-local-llms-on-my-rtx-4070-ti-for-real-work-only-one-earned-a-permanent-spot-jff5bwcpq) · XDA Developers · 1 upvotes · 0 comments - [Mac M3 Max vs RTX 4090: Local LLM Performance Showdown 2026](https://daily.dev/posts/mac-m3-max-vs-rtx-4090-local-llm-performance-showdown-2026-2dh9rvkax) · SitePoint · 1 upvotes · 0 comments - [My 8GB GPU shouldn't run flagship local LLMs, but this workflow makes it work anyway](https://daily.dev/posts/my-8gb-gpu-shouldn-t-run-flagship-local-llms-but-this-workflow-makes-it-work-anyway-nb2r0gq97) · XDA Developers · 1 upvotes · 0 comments --- Tags: [#vibe-coding](https://daily.dev/tags/vibe-coding), [#local-ai](https://daily.dev/tags/local-ai), [#qwen](https://daily.dev/tags/qwen), [#llama-cpp](https://daily.dev/tags/llama-cpp) [View this post on daily.dev](https://daily.dev/posts/zhgzjwota) ```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":"DiscussionForumPosting","mainEntityOfPage":"https://daily.dev/posts/zhgzjwota","headline":"Qwen 3.8 27B: 16GB VRAM Local Test","text":"Shared: Qwen 3.8 27B: 16GB VRAM Local Test","url":"https://daily.dev/posts/zhgzjwota","datePublished":"2026-08-16T03:14:24.226Z","dateModified":"2026-08-16T03:15:15.994Z","author":{"@type":"Person","name":"nerdalytics","url":"https://daily.dev/nerdalytics","image":"https://avatars.githubusercontent.com/u/97166791?v=4","description":"Husband. Cat and dog dad. Nerd.","interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"EndorseAction"},"userInteractionCount":5530}},"image":"https://media.daily.dev/image/upload/s--P4t4XyoV--/f_auto/v1722860399/public/Placeholder%2001","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":25},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":6}],"sharedContent":{"@type":"WebPage","url":"https://api.daily.dev/r/u8lLnnl7u"},"comment":[{"@type":"Comment","text":"I really do not see LLMs of these size to be reasonable for agentic workload.","datePublished":"2026-08-16T05:56:44.164Z","url":"https://daily.dev/posts/ZhgzJwoTA#c-tvGgyAeoj","author":{"@type":"Person","name":"László Szabó","url":"https://daily.dev/lezli01","image":"https://lh3.googleusercontent.com/a/ACg8ocKj3v7EFYJoXUUuro6ALF9fD3RTRATiBpBOVcOzqro4fy6bWqMm=s96-c"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":4}},{"@type":"Comment","text":"Could you rerun the portfolio prompt with a quant that fits entirely in 16 GB of VRAM?","datePublished":"2026-08-16T19:14:16.679Z","url":"https://daily.dev/posts/ZhgzJwoTA#c-cKWvdaXd2","author":{"@type":"Person","name":"Agustin Barrientos","url":"https://daily.dev/agustinbarrientos","image":"https://media.daily.dev/image/upload/s--5ayxQnqn--/f_auto/v1788281802/avatars/avatar_wQYYVe5Tbj0NJ7C7qPoa8?_a=BAMAMicg0"}},{"@type":"Comment","text":"I really loved this thumbnail","datePublished":"2026-08-18T16:13:05.698Z","url":"https://daily.dev/posts/ZhgzJwoTA#c-ged04CrxR","author":{"@type":"Person","name":"Lucas Goularte","url":"https://daily.dev/lucasgoularte","image":"https://lh3.googleusercontent.com/a/ACg8ocI6_xX-FHCNWxitHoNOjKgCRsuK1DWVzdEm4u9pC0Irsf3dT72w=s96-c"}}],"isPartOf":{"@type":"WebPage","url":"https://daily.dev/sources/jcwojcedqzpspqbvk73nc","name":"nerdalytics"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"nerdalytics","item":"https://daily.dev/sources/jcwojcedqzpspqbvk73nc"},{"@type":"ListItem","position":3,"name":"Qwen 3.8 27B: 16GB VRAM Local Test"}]} ```