{"id":3839,"date":"2026-08-16T22:00:00","date_gmt":"2026-08-16T22:00:00","guid":{"rendered":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/"},"modified":"2026-08-16T22:59:28","modified_gmt":"2026-08-16T22:59:28","slug":"qwen-38-27b","status":"publish","type":"post","link":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/","title":{"rendered":"Qwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues"},"content":{"rendered":"


\n<\/p>\n

\n

Qwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues<\/h2>\n

sixteenth August 2026<\/p>\n

Friday\u2019s large launch was Qwen 3.8 27B, an Apache 2 licensed 27B parameter vision-capable LLM from Alibaba\u2019s Qwen analysis lab. I\u2019ve been trying ahead to this one: 27B is a wonderful dimension for operating a mannequin on a fairly specced laptop computer, and its predecessor Qwen 3.6 27B was spectacular.<\/p>\n

Qwen\u2019s self-reported benchmarks for this mannequin are eye-opening. They present a lift from each Qwen 3.6 27B and the closed-weight Qwen 3.7-Plus, which was one among Qwen\u2019s strongest fashions of any dimension as lately as Might this 12 months. It will likely be attention-grabbing to listen to what impartial benchmarks must say concerning the mannequin.<\/p>\n

I\u2019ve been operating the mannequin on two completely different machines: my 128GB M5 Max MacBook Professional, and an NVIDIA DGX Spark. On each machines I\u2019m operating LM Studio and their 17GB Q4_K_M quantized construct. I additionally tried utilizing llama-server instantly on the Spark.<\/p>\n

Qwen\u2019s documentation describes the mannequin as defaulting to xhigh for the reasoning effort, and the LM Studio GGUF I\u2019ve been making an attempt preserves that default:<\/p>\n

\n

Qwen3.8 comes with official assist for reasoning_effort, which can be utilized to regulate reasoning depth and management value:<\/p>\n

xhigh (default): for advanced duties demanding thorough evaluation<\/p>\n

medium: balancing accuracy and velocity<\/p>\n

low: environment friendly reasoning optimizing for velocity and value<\/p>\n<\/blockquote>\n

This can be a hilarious default. It\u2019s completely not a great way to run the mannequin, particularly on client {hardware}. I\u2019ve been discovering the outcomes extraordinarily entertaining.<\/p>\n

I rapidly bumped into issues with LM Studio\u2019s default context restrict of 8,192 tokens\u2014Qwen was utilizing all of them up occupied with even essentially the most mundane of issues. I loaded the mannequin with the total 262,144 most context size and that drawback went away.<\/p>\n

Right here\u2019s the pelican driving a bicycle SVG I bought from my first try with that elevated context size. It took 21 minutes to generate, utilizing 22,276 reasoning tokens to provide 3,223 tokens of output. You’ll be able to learn the reasoning hint right here.<\/p>\n

\"A<\/p>\n

That is by far the very best pelican SVG I\u2019ve been in a position to generate with a mannequin that runs on a neighborhood machine\u2014and this Qwen is fairly small, only a 17GB file on disk. There\u2019s quite a bit to love about this:<\/p>\n

The bicycle body is the appropriate form
\nIt has legs on all sides of the bike\u2014that\u2019s very uncommon
\nGood, clear pelican pouch
\nThe wings prolong to the touch the handlebars!
\nThe movement strains are behind, not in entrance
\nIt has a tasteful background\u2014good solar, clouds, hill, flowers and grass.<\/p>\n

Was that value ready 21 minutes for? Completely not.<\/p>\n

Right here\u2019s that very same immediate run with reasoning turned off\u2014transcript right here. This one produced 3,715 tokens and took 137s\u2014simply over two minutes.<\/p>\n

\"The
\nAnd only for completeness, I used OpenRouter to run the identical immediate via the a lot bigger Qwen 3.8 2.4T-A95B (launched final week) and bought this snazzy animated SVG:<\/p>\n

Your browser doesn’t assist HTML5 video.<\/p>\n

I stated Qwen at xhigh tends to over-think issues, however how dangerous actually is it?<\/p>\n

I attempted a a lot easier immediate, once more with that default further excessive setting:<\/p>\n

\n

draw an svg of a circle<\/p>\n<\/blockquote>\n

Qwen\u2019s reasoning hint began like this:<\/p>\n

\n

The person is asking for an SVG drawing of a circle. Easy request \u2014 however I would like it to be a fastidiously crafted piece. Let me make one thing that goes past simply : a single self-contained SVG file with character \u2014 perhaps a geometrical \u201ccircle examine,\u201d with delicate animation, layered rings, and a particular palette.<\/p>\n

Preserve the scope proper: they requested for an SVG of a circle. So the core is a circle. However I can add craft: concentric information circles (like a compass\/geometry drawing), tick marks, a gentle gradient fill on the principle circle, restrained ambient movement (a slowly rotating dashed ring, pulsing glow). Respect prefers-reduced-motion? That\u2019s extra for CSS; SVG SMIL or CSS inside SVG will do.<\/p>\n

Palette choices: deep teal ink on heat paper? Or daring vermilion circle on off-white with navy development strains \u2014 Bauhaus\/compass-drawing vibe. Let me go along with a \u201cgeometric examine\u201d look: cool slate background, or vivid paper white? Paper white is ok if it\u2019s not the cream-and-terracotta combo. […]<\/p>\n<\/blockquote>\n

A number of minutes later it produced this positively stunning animated circle, which was solely not what I had requested for!<\/p>\n

Your browser doesn’t assist HTML5 video.<\/p>\n

My robust advice: ignore that default. Run Qwen 3.8 27B on low and even no reasoning ranges at first. It\u2019s an excellent mannequin, however wow that default setting is a foul place to start out.<\/p>\n

It\u2019s excellent at bounding packing containers<\/h4>\n

A enjoyable approach to check a imaginative and prescient mannequin is to see how properly it might return bounding packing containers round objects in {a photograph}. I\u2019ve seen earlier Qwen fashions deal properly with this, so I made a decision to place it to the check drawing bounding packing containers round some pelicans.<\/p>\n

I\u2019ve seen asking for 0-1000 scale produce good outcomes up to now. I attempted this:<\/p>\n

llm -a https:\/\/static.inaturalist.org\/images\/714731804\/massive.jpg
\n -m lmstudio\/qwen\/qwen3.8-27b
\n ‘<\/span>Return JSON bounding packing containers for the pelicans on this picture, 0-1000 scale for every dimension‘<\/span><\/span><\/div>\n

Right here\u2019s the reasoning hint, which produced this:<\/p>\n

[
\n {“bbox_2d”<\/span>: [195<\/span>, 290<\/span>, 370<\/span>, 780<\/span>], “label”<\/span>: “<\/span>pelicans“<\/span><\/span>},
\n {“bbox_2d”<\/span>: [445<\/span>, 320<\/span>, 675<\/span>, 850<\/span>], “label”<\/span>: “<\/span>pelicans“<\/span><\/span>}
\n]<\/div>\n

That is such a very good match. Listed below are these packing containers rendered on prime of the picture:<\/p>\n

\"A<\/p>\n

Constructing a device to label bounding packing containers<\/h4>\n

That visualization of the bounding packing containers was taken utilizing a brand new customized device that I had Qwen 3.8 27B construct for me, operating offline on my laptop computer.<\/p>\n

I forgot to dial down the pondering effort so it was massively over-engineered, but it surely did handle to provide this full interface from this single immediate:<\/p>\n

\n[
\n {“bbox_2d”: [195, 290, 370, 780], “label”: “pelicans”},
\n {“bbox_2d”: [445, 320, 675, 850], “label”: “pelicans”}
\n]<\/p>\n

Construct an HTML web page which has an enter field for accepting the URL to a picture and a textarea for accepting the above fashion of JSON.<\/p>\n

It appends the picture to the web page, measures its width and peak, then treats the coords within the bbox_2d as scaled from 0-1000 and scales them in opposition to the precise width and peak, then it renders labelled packing containers over the picture.<\/p>\n<\/blockquote>\n

This screenshot exhibits one of many options I didn’t ask for\u2014a demo scene, for in case you don\u2019t have {a photograph} to check the device with:<\/p>\n

\"Screenshot<\/p>\n

Right here\u2019s the related section of the pondering hint, the place it determined to attract its personal pelicans purely as a result of I had used the label \u201cpelicans\u201d within the instance JSON I gave it within the immediate:<\/p>\n

\n

Additionally a \u201cload pattern\u201d that makes use of a identified picture? Can\u2019t rely on exterior photos, however\u2026 the picture URL enter is user-provided; I may add a \u201cattempt with pattern\u201d button […] Hmm, I can draw a easy scene on canvas, export it as an information URL, and cargo it into the picture \u2014 that\u2019s self-contained and demo-able! […] However the person\u2019s coords are for an precise pelican picture; a generated placeholder can nonetheless demo the scaling. Generate a 1000×1000 placeholder: gradient water + two blob-like \u201cpelican\u201d silhouettes positioned on the given bboxes (utilizing the identical scale \u2014 cute: silhouettes on the precise 0-1000 positions, displaying the packing containers align). This makes for a enjoyable, self-contained demo. Preserve it easy: sky gradient, solar, water, two pelican-ish shapes (ellipse physique, circle head, beak). Place at bbox facilities.<\/p>\n<\/blockquote>\n

(I\u2019m barely nervous that fashions world wide might need a bias in the direction of drawing pelicans at any probability they’ll get, introduced on by almost two years of publicity to my very own silly benchmark.)<\/p>\n

Is all that over-thinking essential? Perhaps it’s, no less than a bit. I attempted with reasoning turned off and bought this model, (transcript right here), which almost works however exhibits the packing containers within the fallacious place:<\/p>\n

\"BBox<\/p>\n

So with out reasoning it didn\u2019t fairly one-shot a working device. I\u2019m positive it may get there with some follow-up prompts, however it is a good instance of how reasoning could make a distinction.<\/p>\n

Sure, it might drive coding brokers<\/h4>\n

One of many largest questions round native fashions is whether or not or not they’ve sufficient horsepower to efficiently run a coding agent loop. Coding brokers require lengthy context, robust code technology assist and dependable tool-calling. On paper Qwen 3.8 27B has all three of those, so is it as much as the duty?<\/p>\n

My preliminary experiments with Pi have been very promising. I selected Pi as a result of it has a shorter system immediate than most different choices, making it a greater match for making an attempt out smaller fashions.<\/p>\n

I configured Pi to make use of Qwen 3.8 27B operating in LM Studio on the Spark (shared by way of tailscale serve) by including this to ~\/.pi\/agent\/fashions.json:<\/p>\n

{
\n “suppliers”<\/span>: {
\n “spark”<\/span>: {
\n “baseUrl”<\/span>: “<\/span>https:\/\/spark-18b3.tail68a31.ts.web\/v1“<\/span><\/span>,
\n “api”<\/span>: “<\/span>openai-responses“<\/span><\/span>,
\n “apiKey”<\/span>: “<\/span>dummy“<\/span><\/span>,
\n “fashions”<\/span>: [
\n {
\n “id”<\/span>: “<\/span>qwen3.8-27b“<\/span><\/span>,
\n “reasoning”<\/span>: true<\/span>
\n }
\n ]
\n }
\n }
\n}<\/div>\n

Then ran pi –provider spark –model qwen3.8-27b in my ~\/dev\/datasette folder and prompted:<\/p>\n

\n

how does auth work?<\/p>\n<\/blockquote>\n

After a sequence of reasoning and power calls that accessed a bunch of various recordsdata it produced this reply, which may be very stable.<\/p>\n

Only one drawback: I wished to share that transcript. So I pointed Pi and Qwen 3.8 27B on the JSONL transcript file in ~\/.pi\/agent\/classes\/–Customers-simon-Dropbox-dev-datasette– and prompted:<\/p>\n

\n

Write Python code to transform this jsonl to markdown<\/p>\n<\/blockquote>\n

And it constructed and examined this pi_jsonl_to_md.py, which did precisely what I wanted. Right here\u2019s that session transcript, printed utilizing the device that it created.<\/p>\n

The hunt for velocity<\/h4>\n

Thus far that is all trying very promising. We now have a 17GB mannequin that runs on high-end client {hardware} and may write code, drive instruments, annotate photos and customarily do all the things that I would like from an LLM for getting actual work completed.<\/p>\n

There\u2019s one very important catch: it feels gradual\u2014particularly when it begins over-thinking, however even with out that it\u2019s not significantly sprightly.<\/p>\n

I\u2019ve been getting round 15-30 tokens a second from LM Studio. That\u2019s not horrible, but it surely\u2019s gradual sufficient that it\u2019s going to be laborious to win me away from hosted API fashions, which might return outcomes a complete lot quicker. Synthetic Evaluation observe token velocity and present OpenAI 5.6 Sol at 74 tokens\/second and 5.6 Luna at a formidable 184\/second.<\/p>\n

The excellent news is that the group have been exploring methods to hurry issues up for the reason that mannequin was first launched two days in the past.<\/p>\n

One of the promising optimizations is baked into the mannequin itself. Qwen helps Multi-Token Prediction, an structure trick the place a less expensive mechanism guesses a number of tokens forward and the principle mannequin can then rapidly confirm if the guesses have been appropriate. This could have fairly a dramatic impact on inference efficiency.<\/p>\n

Primarily based on this tweet from llama.cpp creator Georgi Gerganov I attempted operating the mannequin with MTP like this on the Spark:<\/p>\n

llama serve
\n -hf ggml-org\/Qwen3.8-27B-GGUF:Q4_K_M
\n -hfd ggml-org\/Qwen3.8-27B-GGUF:Q4_0
\n –spec-default
\n –spec-type draft-mtp
\n –reasoning-preserve<\/div>\n

And positive sufficient, this gave me a big increase. I had GPT-5.6 in Codex run a comparative benchmark on the Spark and the –spec-type draft-mtp server outperformed the LM Studio default GGUF by round 72%.<\/p>\n

I count on we\u2019ll see a complete lot extra innovation round serving this mannequin quicker over the following few weeks. The MLX group doubtless have some tips brewing as properly.<\/p>\n

Some observations<\/h4>\n

The truth that a 17GB file can do all of these things on my dwelling machines is a miracle. As soon as once more, I\u2019m delighted and amazed at how a lot progress native fashions have made this 12 months. A 12 months in the past this is able to have been aggressive with the very best and costliest of the proprietary fashions\u2014at this time it might run on a succesful laptop computer.<\/p>\n

The one factor holding this again from being a every day driver is efficiency. It feels fairly gradual on each the M5 Mac and the DGX Spark. That\u2019s the catch with these dense (non-Combination-of-Consultants) fashions\u2014they require a complete lot of reminiscence bandwidth to carry out properly, and neither of the machines I’ve entry to are prime performers in that regard.<\/p>\n

Crucial factor about Qwen 3.8 27B is what it demonstrates. We will have an open weights basic objective mannequin with an extended context, efficient device calling, robust imaginative and prescient capacity, and competent code technology, and we are able to match the entire thing in only a 17GB file.<\/p>\n

The fashions at this dimension proceed to get higher at a formidable fee. We don\u2019t have to spend half one million {dollars} on datacenter-class {hardware} simply to run a reliable mannequin.<\/p>\n<\/div>\n


\n
Source link <\/a><\/p>\n","protected":false},"excerpt":{"rendered":"

Qwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues sixteenth August 2026 Friday\u2019s large launch was Qwen 3.8 27B, an Apache 2 licensed 27B parameter vision-capable LLM from Alibaba\u2019s Qwen analysis lab. I\u2019ve been trying ahead to this one: 27B is a wonderful dimension for operating a mannequin on a fairly […]<\/p>\n","protected":false},"author":1,"featured_media":3841,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"fifu_image_url":"https:\/\/static.simonwillison.net\/static\/2026\/qwen-thinking-bicycle-27b.jpg","fifu_image_alt":"","jnews-multi-image_gallery":[],"jnews_single_post":[],"jnews_primary_category":[],"jnews_override_bookmark_settings":[],"jnews_social_meta":[],"jnews_override_counter":[],"footnotes":""},"categories":[2],"tags":[4159,4161,4160,4163,4158,4162],"class_list":["post-3839","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-research-breakthroughs","tag-27b","tag-defaults","tag-excellent","tag-overthinking","tag-qwen","tag-wildly"],"yoast_head":"\nQwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues - Future News 24<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Qwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues - Future News 24\" \/>\n<meta property=\"og:description\" content=\"Qwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues sixteenth August 2026 Friday\u2019s large launch was Qwen 3.8 27B, an Apache 2 licensed 27B parameter vision-capable LLM from Alibaba\u2019s Qwen analysis lab. I\u2019ve been trying ahead to this one: 27B is a wonderful dimension for operating a mannequin on a fairly […]\" \/>\n<meta property=\"og:url\" content=\"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/\" \/>\n<meta property=\"og:site_name\" content=\"Future News 24\" \/>\n<meta property=\"article:published_time\" content=\"2026-08-16T22:00:00+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2026-08-16T22:59:28+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/static.simonwillison.net\/static\/2026\/qwen-thinking-bicycle-27b.jpg\" \/>\n<meta name=\"author\" content=\"Future News 24\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:image\" content=\"https:\/\/static.simonwillison.net\/static\/2026\/qwen-thinking-bicycle-27b.jpg\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Future News 24\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"11 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/08\\\/16\\\/qwen-38-27b\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/08\\\/16\\\/qwen-38-27b\\\/\"},\"author\":{\"name\":\"Future News 24\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#\\\/schema\\\/person\\\/cecad1bde21cfc357cf70128144d6c83\"},\"headline\":\"Qwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues\",\"datePublished\":\"2026-08-16T22:00:00+00:00\",\"dateModified\":\"2026-08-16T22:59:28+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/08\\\/16\\\/qwen-38-27b\\\/\"},\"wordCount\":2221,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/08\\\/16\\\/qwen-38-27b\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/static.simonwillison.net\\\/static\\\/2026\\\/qwen-thinking-bicycle-27b.jpg\",\"keywords\":[\"27B\",\"defaults\",\"excellent\",\"overthinking\",\"Qwen\",\"wildly\"],\"articleSection\":[\"AI Research & Breakthroughs\"],\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/08\\\/16\\\/qwen-38-27b\\\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/08\\\/16\\\/qwen-38-27b\\\/\",\"url\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/08\\\/16\\\/qwen-38-27b\\\/\",\"name\":\"Qwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues - Future News 24\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/08\\\/16\\\/qwen-38-27b\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/08\\\/16\\\/qwen-38-27b\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/static.simonwillison.net\\\/static\\\/2026\\\/qwen-thinking-bicycle-27b.jpg\",\"datePublished\":\"2026-08-16T22:00:00+00:00\",\"dateModified\":\"2026-08-16T22:59:28+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/08\\\/16\\\/qwen-38-27b\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/08\\\/16\\\/qwen-38-27b\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/08\\\/16\\\/qwen-38-27b\\\/#primaryimage\",\"url\":\"https:\\\/\\\/static.simonwillison.net\\\/static\\\/2026\\\/qwen-thinking-bicycle-27b.jpg\",\"contentUrl\":\"https:\\\/\\\/static.simonwillison.net\\\/static\\\/2026\\\/qwen-thinking-bicycle-27b.jpg\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/2026\\\/08\\\/16\\\/qwen-38-27b\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/futurenews24.com\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Qwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#website\",\"url\":\"https:\\\/\\\/futurenews24.com\\\/\",\"name\":\"Future News 24\",\"description\":\"The Smart Hub for AI and Next-Gen Innovation\",\"publisher\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/futurenews24.com\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#organization\",\"name\":\"Future News 24\",\"url\":\"https:\\\/\\\/futurenews24.com\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/futurenews24.com\\\/wp-content\\\/uploads\\\/2026\\\/06\\\/fn24-favicon.png\",\"contentUrl\":\"https:\\\/\\\/futurenews24.com\\\/wp-content\\\/uploads\\\/2026\\\/06\\\/fn24-favicon.png\",\"width\":250,\"height\":250,\"caption\":\"Future News 24\"},\"image\":{\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#\\\/schema\\\/logo\\\/image\\\/\"}},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/futurenews24.com\\\/#\\\/schema\\\/person\\\/cecad1bde21cfc357cf70128144d6c83\",\"name\":\"Future News 24\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/d57f07142d73cb5503ab2446ea7bc9ef3d0a5ba378d64a6157692311e42bf097?s=96&d=mm&r=g\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/d57f07142d73cb5503ab2446ea7bc9ef3d0a5ba378d64a6157692311e42bf097?s=96&d=mm&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/d57f07142d73cb5503ab2446ea7bc9ef3d0a5ba378d64a6157692311e42bf097?s=96&d=mm&r=g\",\"caption\":\"Future News 24\"},\"sameAs\":[\"https:\\\/\\\/futurenews24.com\"],\"url\":\"https:\\\/\\\/futurenews24.com\\\/index.php\\\/author\\\/mridulpahuja20\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Qwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues - Future News 24","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/","og_locale":"en_US","og_type":"article","og_title":"Qwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues - Future News 24","og_description":"Qwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues sixteenth August 2026 Friday\u2019s large launch was Qwen 3.8 27B, an Apache 2 licensed 27B parameter vision-capable LLM from Alibaba\u2019s Qwen analysis lab. I\u2019ve been trying ahead to this one: 27B is a wonderful dimension for operating a mannequin on a fairly […]","og_url":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/","og_site_name":"Future News 24","article_published_time":"2026-08-16T22:00:00+00:00","article_modified_time":"2026-08-16T22:59:28+00:00","og_image":[{"url":"https:\/\/static.simonwillison.net\/static\/2026\/qwen-thinking-bicycle-27b.jpg","type":"","width":"","height":""}],"author":"Future News 24","twitter_card":"summary_large_image","twitter_image":"https:\/\/static.simonwillison.net\/static\/2026\/qwen-thinking-bicycle-27b.jpg","twitter_misc":{"Written by":"Future News 24","Est. reading time":"11 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/#article","isPartOf":{"@id":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/"},"author":{"name":"Future News 24","@id":"https:\/\/futurenews24.com\/#\/schema\/person\/cecad1bde21cfc357cf70128144d6c83"},"headline":"Qwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues","datePublished":"2026-08-16T22:00:00+00:00","dateModified":"2026-08-16T22:59:28+00:00","mainEntityOfPage":{"@id":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/"},"wordCount":2221,"commentCount":0,"publisher":{"@id":"https:\/\/futurenews24.com\/#organization"},"image":{"@id":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/#primaryimage"},"thumbnailUrl":"https:\/\/static.simonwillison.net\/static\/2026\/qwen-thinking-bicycle-27b.jpg","keywords":["27B","defaults","excellent","overthinking","Qwen","wildly"],"articleSection":["AI Research & Breakthroughs"],"inLanguage":"en-US","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/","url":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/","name":"Qwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues - Future News 24","isPartOf":{"@id":"https:\/\/futurenews24.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/#primaryimage"},"image":{"@id":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/#primaryimage"},"thumbnailUrl":"https:\/\/static.simonwillison.net\/static\/2026\/qwen-thinking-bicycle-27b.jpg","datePublished":"2026-08-16T22:00:00+00:00","dateModified":"2026-08-16T22:59:28+00:00","breadcrumb":{"@id":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/#primaryimage","url":"https:\/\/static.simonwillison.net\/static\/2026\/qwen-thinking-bicycle-27b.jpg","contentUrl":"https:\/\/static.simonwillison.net\/static\/2026\/qwen-thinking-bicycle-27b.jpg"},{"@type":"BreadcrumbList","@id":"https:\/\/futurenews24.com\/index.php\/2026\/08\/16\/qwen-38-27b\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/futurenews24.com\/"},{"@type":"ListItem","position":2,"name":"Qwen 3.8 27B is superb, but it surely defaults to wildly overthinking issues"}]},{"@type":"WebSite","@id":"https:\/\/futurenews24.com\/#website","url":"https:\/\/futurenews24.com\/","name":"Future News 24","description":"The Smart Hub for AI and Next-Gen Innovation","publisher":{"@id":"https:\/\/futurenews24.com\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/futurenews24.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/futurenews24.com\/#organization","name":"Future News 24","url":"https:\/\/futurenews24.com\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/futurenews24.com\/#\/schema\/logo\/image\/","url":"https:\/\/futurenews24.com\/wp-content\/uploads\/2026\/06\/fn24-favicon.png","contentUrl":"https:\/\/futurenews24.com\/wp-content\/uploads\/2026\/06\/fn24-favicon.png","width":250,"height":250,"caption":"Future News 24"},"image":{"@id":"https:\/\/futurenews24.com\/#\/schema\/logo\/image\/"}},{"@type":"Person","@id":"https:\/\/futurenews24.com\/#\/schema\/person\/cecad1bde21cfc357cf70128144d6c83","name":"Future News 24","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/secure.gravatar.com\/avatar\/d57f07142d73cb5503ab2446ea7bc9ef3d0a5ba378d64a6157692311e42bf097?s=96&d=mm&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/d57f07142d73cb5503ab2446ea7bc9ef3d0a5ba378d64a6157692311e42bf097?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/d57f07142d73cb5503ab2446ea7bc9ef3d0a5ba378d64a6157692311e42bf097?s=96&d=mm&r=g","caption":"Future News 24"},"sameAs":["https:\/\/futurenews24.com"],"url":"https:\/\/futurenews24.com\/index.php\/author\/mridulpahuja20\/"}]}},"_links":{"self":[{"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/posts\/3839","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/comments?post=3839"}],"version-history":[{"count":1,"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/posts\/3839\/revisions"}],"predecessor-version":[{"id":3840,"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/posts\/3839\/revisions\/3840"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/media\/3841"}],"wp:attachment":[{"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/media?parent=3839"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/categories?post=3839"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/futurenews24.com\/index.php\/wp-json\/wp\/v2\/tags?post=3839"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}