{"id":13811,"date":"2026-11-17T11:02:46","date_gmt":"2026-11-17T02:02:46","guid":{"rendered":"https:\/\/localmodelwatch.tsuchitsuchi.com\/?p=13811"},"modified":"2026-11-17T11:11:57","modified_gmt":"2026-11-17T02:11:57","slug":"ollama-v04012-released","status":"publish","type":"post","link":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/ollama-v04012-released\/","title":{"rendered":"Ollama v0.40.12 Released: Adds Liquid AI LFM3 and d1 Models"},"content":{"rendered":"<p><!-- lmw:facts --><\/p>\n<h2>At a Glance<\/h2>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>Item<\/th>\n<th>Value<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Repository<\/td>\n<td><a href=\"https:\/\/github.com\/ollama\/ollama\">ollama\/ollama<\/a><\/td>\n<\/tr>\n<tr>\n<td>Version<\/td>\n<td><a href=\"https:\/\/github.com\/ollama\/ollama\/releases\/tag\/v0.40.12\">v0.40.12<\/a><\/td>\n<\/tr>\n<tr>\n<td>Published<\/td>\n<td>2026-10-27<\/td>\n<\/tr>\n<tr>\n<td>License<\/td>\n<td>MIT<\/td>\n<\/tr>\n<tr>\n<td>Source type<\/td>\n<td>Primary source (the publisher itself)<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><em>Values determined by this site&#8217;s code when the information was collected. Dates are JST.<\/em><\/p>\n<p><!-- \/lmw:facts --><\/p>\n<h2>Overview<\/h2>\n<p>Ollama has released its latest version, v0.40.12. Ollama is a tool designed to easily run various open-weight models, such as Gemma, DeepSeek, and Qwen, in local environments. This update adds support for Liquid AI&#8217;s new model family, &#8220;LFM3&#8221; and &#8220;d1 decision models,&#8221; and brings faster prompt processing for specific models like Gemma 4, particularly in Apple Silicon environments.<\/p>\n<h2>Breaking Changes and Deprecations<\/h2>\n<p>None.<\/p>\n<h2>Key Changes<\/h2>\n<h3>Support for the Liquid AI LFM3 Family<\/h3>\n<p>Support has been added for the &#8220;LFM3&#8221; model family developed by Liquid AI, designed for local agents, tool calling, and structured outputs. This family includes <code>lfm3<\/code>, a 1.7B text model with a 256K token context window; <code>lfm3-vl<\/code>, which adds image understanding; <code>lfm3-audio<\/code>, capable of speech recognition, translation, and text-to-speech across 18 languages; and <code>lfm3-omni<\/code>, which handles text, images, and audio in a single model.<\/p>\n<p>Note that while this release supports running <code>lfm3<\/code> and <code>d1 2b<\/code> on Apple hardware, support for Linux and Windows is scheduled to arrive soon.<\/p>\n<h3>Addition of d1 Decision Models (2B)<\/h3>\n<p>The 2B size of Liquid AI&#8217;s <code>d1<\/code> model, built on the LFM3 architecture, has been added. <code>d1<\/code> makes decisions regarding text and images, while <code>d1-omni<\/code> also supports audio. For example, it can be run using a command like the following:<\/p>\n<pre><code class=\"language-sh\">ollama run d1:2b &quot;I was charged twice for order 4417.&quot; --ask &quot;Is this a billing issue?&quot;\n<\/code><\/pre>\n<h3>Faster Prompt Processing on Apple Silicon<\/h3>\n<p>For users running Apple Silicon, prompt processing speeds have been improved for the Gemma 4, Flint, and Laguna models. Users running these models on Apple chips such as Macs can expect smoother operation following this update.<\/p>\n<h3>Inference Improvements and Bug Fixes<\/h3>\n<p>Improvements have been made to inference performance and behavior in specific environments. DFlash2 draft models are now supported for speculative decoding in MLX environments. Additionally, Gemma 4&#8217;s ability to retain detailed information contained in small images has been improved. Furthermore, <code>ollama launch<\/code> now works with OpenCode v2, and fixes related to tool-only assistant turns in Kimi CLI are also included. These changes accompany internal updates to MLX and llama.cpp.<\/p>\n<h2>Supported Models and Hardware<\/h2>\n<p>With this update, the following model architectures and models are now available:<\/p>\n<h3>Liquid AI LFM3 Series<\/h3>\n<p>The LFM3 family, based on a new &#8220;tri-hybrid architecture&#8221; developed by Liquid AI, is supported. These models are designed for local agents, tool calling, and structured outputs.<\/p>\n<ul>\n<li><code>lfm3<\/code>: A text model with 1.7B parameters. It supports a 256K token context window.<\/li>\n<li><code>lfm3-vl<\/code>: A model adding image understanding capabilities to <code>lfm3<\/code>.<\/li>\n<li><code>lfm3-audio<\/code>: A model supporting speech recognition, translation, and text-to-speech in 18 languages.<\/li>\n<li><code>lfm3-omni<\/code>: A multimodal model capable of handling text, images, and audio in an integrated manner within a single model.<\/li>\n<\/ul>\n<h3>d1 Decision Models<\/h3>\n<p>The 2B size of the <code>d1<\/code> model, specialized for decision making and built on the LFM3 architecture, has been added.<\/p>\n<ul>\n<li><code>d1<\/code>: A model that makes decisions regarding text and images.<\/li>\n<li><code>d1-omni<\/code>: A model adding audio capabilities to <code>d1<\/code>.<\/li>\n<\/ul>\n<h3>Supported Hardware and Optimizations<\/h3>\n<p>Regarding hardware and runtime environments, the following points are reported:<\/p>\n<ul>\n<li><strong>Apple Hardware<\/strong>: Running <code>lfm3<\/code> and <code>d1 2b<\/code> is supported on Apple hardware in this release. In addition, prompt processing for Gemma 4, Flint, and Laguna models has been accelerated in Apple Silicon environments. Furthermore, Gemma 4 has improved capability in retaining detailed information within small images.<\/li>\n<li><strong>MLX Environments<\/strong>: DFlash2 draft models are newly supported in speculative decoding.<\/li>\n<li><strong>Other Platforms<\/strong>: Support for <code>lfm3<\/code> and <code>d1 2b<\/code> on Linux and Windows environments is reported to be coming soon.<\/li>\n<\/ul>\n<h2>How to Get It<\/h2>\n<p>To update Ollama to the latest version, obtain and apply the latest installer from the official releases page.<\/p>\n<p>When using new models, use the standard <code>ollama run<\/code> command. For example, to run <code>lfm3<\/code>:<\/p>\n<pre><code class=\"language-sh\">ollama run lfm3\n<\/code><\/pre>\n<p>Additionally, to run instructions combining text and questions using the <code>d1:2b<\/code> model, you can execute it in the following format:<\/p>\n<pre><code class=\"language-sh\">ollama run d1:2b &quot;I was charged twice for order 4417.&quot; --ask &quot;Is this a billing issue?&quot;\n<\/code><\/pre>\n<p><!-- lmw:releases --><\/p>\n<h2>Releases Since Our Last Article<\/h2>\n<p><em>Compiled by Local Model Watch from the project&#8217;s GitHub releases: the versions between this release and the last one we covered, which did not get separate articles.<\/em> <em>Full history: <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/engine-ollama-en\/\">release tracker<\/a>.<\/em><\/p>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>Version<\/th>\n<th>Released<\/th>\n<th>Release notes<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>v0.40.11<\/td>\n<td>2026-10-24<\/td>\n<td><a href=\"https:\/\/github.com\/ollama\/ollama\/releases\/tag\/v0.40.11\">GitHub<\/a><\/td>\n<\/tr>\n<tr>\n<td>v0.40.10<\/td>\n<td>2026-10-24<\/td>\n<td><a href=\"https:\/\/github.com\/ollama\/ollama\/releases\/tag\/v0.40.10\">GitHub<\/a><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><!-- \/lmw:releases --><\/p>\n<p><!-- lmw:related --><\/p>\n<h2>Related Articles<\/h2>\n<ul>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/ollama-v0-40-9-released-2\/\">Ollama v0.40.9 Released with Apple Silicon Phonon Speech Support<\/a><\/li>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/ollama-v0-40-4-released\/\">Ollama v0.40.4 Released with Decision Models Support<\/a><\/li>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/ollama-v0-40-15-released\/\">Ollama v0.40.15 Released with Parallel Vision GGUF Support<\/a><\/li>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/ollama-v0-40-5-released\/\">Ollama v0.40.5 Released: NVIDIA Nemotron 3.5 Support Added<\/a><\/li>\n<\/ul>\n<p><!-- \/lmw:related --><\/p>\n<p><!-- lmw:next-steps --><\/p>\n<h2>What to Read Next<\/h2>\n<ul>\n<li><strong>Follow this tool<\/strong> \u2192 <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/engine-ollama-en\/\">Ollama overview and release history (200 releases tracked)<\/a><\/li>\n<li><strong>Other inference engines and runtimes<\/strong> \u2192 <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/engine-llama-cpp-en\/\">llama.cpp<\/a> \/ <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/engine-vllm-en\/\">vLLM<\/a> \/ <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/engine-sglang-en\/\">SGLang<\/a><\/li>\n<\/ul>\n<p><!-- \/lmw:next-steps --><\/p>\n<h2>Sources<\/h2>\n<ul>\n<li><a href=\"https:\/\/github.com\/ollama\/ollama\/releases\/tag\/v0.40.12\">https:\/\/github.com\/ollama\/ollama\/releases\/tag\/v0.40.12<\/a><\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"<p>Ollama v0.40.12 is out, adding support for Liquid AI LFM3 and d1 decision models, plus Apple Silicon prompt processing speedups.<\/p>\n","protected":false},"author":1,"featured_media":13810,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[105],"tags":[505,4242,1032,4177,3282,1093,1547],"class_list":["post-13811","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-engines-and-tools","tag-apple-silicon-en","tag-d1-en","tag-gemma-en","tag-lfm3-en","tag-liquid-ai-en","tag-ollama-en","tag-verified"],"lang":"en","translations":{"en":13811,"ja":13809},"_links":{"self":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts\/13811","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/comments?post=13811"}],"version-history":[{"count":2,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts\/13811\/revisions"}],"predecessor-version":[{"id":17970,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts\/13811\/revisions\/17970"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/media\/13810"}],"wp:attachment":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/media?parent=13811"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/categories?post=13811"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/tags?post=13811"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}