--- title: My local LLM setup got faster after I stopped obsessing... description: Running large LLMs on consumer hardware often means poor performance due to VRAM constraints and RAM offloading. Parameter count is a misleading metric — a 20B... canonical: https://daily.dev/posts/my-local-llm-setup-got-faster-after-i-stopped-obsessing-over-parameter-count-py8knxbbm twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: My local LLM setup got faster after I stopped obsessing over parameter count | daily.dev og:description: Running large LLMs on consumer hardware often means poor performance due to VRAM constraints and RAM offloading. Parameter count is a misleading metric — a 20B... og:url: https://daily.dev/posts/my-local-llm-setup-got-faster-after-i-stopped-obsessing-over-parameter-count-py8knxbbm og:image: https://api.daily.dev/og/posts/py8kNXBBM.png og:image:alt: My local LLM setup got faster after I stopped obsessing over parameter count 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. # My local LLM setup got faster after I stopped obsessing over parameter count **[XDA Developers](https://daily.dev/sources/xda-developers)** · 4 min read · 0 upvotes · 0 comments ## Summary Running large LLMs on consumer hardware often means poor performance due to VRAM constraints and RAM offloading. Parameter count is a misleading metric — a 20B model running with offloading can drop from 40 tokens/sec to 8. More useful factors include architecture (e.g., hybrid Mamba-transformer models like Granite 4 and LFM2 that cut memory use by up to 70%), quantization, and VRAM fit. A better approach is building a small stable of specialized models: a tiny routing model under 1B, a multimodal model loaded on demand, and a reasoning model sized to what your hardware can actually run properly. The under-4B lane handles most tasks well on consumer hardware. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Similar posts on daily.dev - [I replaced my local LLM with a model half its size and got better results — And it wasn't about the parameters](https://daily.dev/posts/i-replaced-my-local-llm-with-a-model-half-its-size-and-got-better-results-and-it-wasn-t-about-the--09vxd04go) · XDA Developers · 0 upvotes · 0 comments - [How much VRAM do you need to run an LLM locally?](https://daily.dev/posts/how-much-vram-do-you-need-to-run-an-llm-locally--sjxd57zfc) · Flavio Copes · 2 upvotes · 1 comments - [Optimizing Local LLMs for Low-End Hardware: 8GB GPU Guide](https://daily.dev/posts/optimizing-local-llms-for-low-end-hardware-8gb-gpu-guide-h3kyjc7i7) · SitePoint · 0 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: [#local-ai](https://daily.dev/tags/local-ai), [#ollama](https://daily.dev/tags/ollama) [View this post on daily.dev](https://daily.dev/posts/my-local-llm-setup-got-faster-after-i-stopped-obsessing-over-parameter-count-py8knxbbm) ```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":"TechArticle","headline":"My local LLM setup got faster after I stopped obsessing over parameter count","url":"https://daily.dev/posts/my-local-llm-setup-got-faster-after-i-stopped-obsessing-over-parameter-count-py8knxbbm","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/my-local-llm-setup-got-faster-after-i-stopped-obsessing-over-parameter-count-py8knxbbm"},"datePublished":"2026-07-23T17:02:33.342Z","dateModified":"2026-07-23T18:26:11.209Z","description":"Running large LLMs on consumer hardware often means poor performance due to VRAM constraints and RAM offloading. Parameter count is a misleading metric — a 20B...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/fcfcd9d38e6590163aff1963fcb61322?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/fcfcd9d38e6590163aff1963fcb61322?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"XDA Developers","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"XDA Developers","logo":"https://media.daily.dev/image/upload/s--kCg6yyAP--/f_auto,q_auto/v1774964407/logos/xda-developers?_a=BAMAMiWQ0","url":"https://daily.dev/sources/xda-developers"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/my-local-llm-setup-got-faster-after-i-stopped-obsessing-over-parameter-count-py8knxbbm","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"local-ai,ollama","timeRequired":"PT4M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"XDA Developers","item":"https://daily.dev/sources/xda-developers"},{"@type":"ListItem","position":3,"name":"My local LLM setup got faster after I stopped obsessing over parameter count"}]} ```