--- title: Mesh LLM: distributed AI computing on iroh | daily.dev description: Mesh LLM pools GPU resources across multiple machines into a single OpenAI-compatible API endpoint, built on top of iroh — a peer-to-peer networking library.... canonical: https://daily.dev/posts/mesh-llm-distributed-ai-computing-on-iroh-ymmyrv7xv twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Mesh LLM: distributed AI computing on iroh | daily.dev og:description: Mesh LLM pools GPU resources across multiple machines into a single OpenAI-compatible API endpoint, built on top of iroh — a peer-to-peer networking library.... og:url: https://daily.dev/posts/mesh-llm-distributed-ai-computing-on-iroh-ymmyrv7xv og:image: https://api.daily.dev/og/posts/ymMyRV7xV.png og:image:alt: Mesh LLM: distributed AI computing on iroh 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. # Mesh LLM: distributed AI computing on iroh **[Hacker News](https://daily.dev/sources/hn)** · 5 min read · 3 upvotes · 0 comments ## Summary Mesh LLM pools GPU resources across multiple machines into a single OpenAI-compatible API endpoint, built on top of iroh — a peer-to-peer networking library. Each node boots an iroh endpoint (identified by public key) that handles NAT traversal and authenticated QUIC connections without a central server. Models can run locally, be routed to a peer that has them loaded, or be split across multiple machines in a pipeline (called 'Skippy') for models too large for any single GPU. The protocol uses QUIC ALPN negotiation with three distinct protocols for mesh communication, control plane, and latency-sensitive activation transport. The system exposes itself as localhost:9337/v1 to any standard OpenAI client, hiding all the distributed complexity. It ships with 40+ models ranging from sub-billion parameter models to 235B MoE giants. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How does Mesh LLM split a large language model across multiple machines that individually can't hold it? Mesh LLM partitions the model by layer ranges into pipeline stages, for example layers 0 to 15 on one node and 16 to 31 on the next, with activations flowing between stages over a dedicated skippy-stage/2 QUIC ALPN channel optimized for low latency. This split mode, internally called 'Skippy,' lets several modest machines together run a model none of them could hold alone, while the OpenAI client still only talks to localhost. _daily.dev surfaces distributed-inference approaches like this for engineers weighing self-hosted versus API-based LLM serving._ ### How does iroh handle networking between nodes in a peer-to-peer mesh without a central server? Every node boots an iroh endpoint identified by a public key, and iroh handles hole-punching, NAT traversal, and relay fallback to open a direct, authenticated QUIC connection between any two nodes regardless of network location. There is no central server; Mesh LLM adds its own gossip layer on top of iroh's transport to control mesh admission, version compatibility, and peer trust. _Developers evaluating peer-to-peer transport layers for distributed apps can track approaches like this on daily.dev._ ## Similar posts on daily.dev - [I put my local LLM on Tailscale, and now I can use it from the other side of the world](https://daily.dev/posts/i-put-my-local-llm-on-tailscale-and-now-i-can-use-it-from-the-other-side-of-the-world-yuoqayjeg) · XDA Developers · 2 upvotes · 2 comments - [Microsoft Three-Layer LLM Routing Architecture for AI Agents on AKS](https://daily.dev/posts/microsoft-three-layer-llm-routing-architecture-for-ai-agents-on-aks-67wgdmok9) · InfoQ · 1 upvotes · 0 comments - [Guide to Local LLMs in 2026: Privacy, Tools & Hardware](https://daily.dev/posts/guide-to-local-llms-in-2026-privacy-tools-hardware-yigh17bqi) · SitePoint · 1 upvotes · 0 comments - [Model-as-a-Service: How to run your own private AI API](https://daily.dev/posts/model-as-a-service-how-to-run-your-own-private-ai-api-pwly2v2wj) · Red Hat Developer · 0 upvotes · 0 comments - [Understanding disaggregated GenAI model serving with llm-d](https://daily.dev/posts/understanding-disaggregated-genai-model-serving-with-llm-d-lvuvnpaha) · Ubuntu · 1 upvotes · 0 comments --- Tags: [#llm](https://daily.dev/tags/llm), [#openai](https://daily.dev/tags/openai), [#distributed-systems](https://daily.dev/tags/distributed-systems) [View this post on daily.dev](https://daily.dev/posts/mesh-llm-distributed-ai-computing-on-iroh-ymmyrv7xv) ```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":"Mesh LLM: distributed AI computing on iroh","url":"https://daily.dev/posts/mesh-llm-distributed-ai-computing-on-iroh-ymmyrv7xv","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/mesh-llm-distributed-ai-computing-on-iroh-ymmyrv7xv"},"datePublished":"2026-07-12T00:07:46.273Z","dateModified":"2026-09-13T18:50:23.000Z","description":"Mesh LLM pools GPU resources across multiple machines into a single OpenAI-compatible API endpoint, built on top of iroh — a peer-to-peer networking library....","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/6da99bed1fac1c789c9f8f8e63d20a16?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/6da99bed1fac1c789c9f8f8e63d20a16?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Hacker News","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":"Hacker News","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/hn","url":"https://daily.dev/sources/hn"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/mesh-llm-distributed-ai-computing-on-iroh-ymmyrv7xv","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":3},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"llm,openai,distributed-systems","timeRequired":"PT5M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Hacker News","item":"https://daily.dev/sources/hn"},{"@type":"ListItem","position":3,"name":"Mesh LLM: distributed AI computing on iroh"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/mesh-llm-distributed-ai-computing-on-iroh-ymmyrv7xv#faq","mainEntity":[{"@type":"Question","name":"How does Mesh LLM split a large language model across multiple machines that individually can't hold it?","acceptedAnswer":{"@type":"Answer","text":"Mesh LLM partitions the model by layer ranges into pipeline stages, for example layers 0 to 15 on one node and 16 to 31 on the next, with activations flowing between stages over a dedicated skippy-stage/2 QUIC ALPN channel optimized for low latency. This split mode, internally called 'Skippy,' lets several modest machines together run a model none of them could hold alone, while the OpenAI client still only talks to localhost. daily.dev surfaces distributed-inference approaches like this for engineers weighing self-hosted versus API-based LLM serving."}},{"@type":"Question","name":"How does iroh handle networking between nodes in a peer-to-peer mesh without a central server?","acceptedAnswer":{"@type":"Answer","text":"Every node boots an iroh endpoint identified by a public key, and iroh handles hole-punching, NAT traversal, and relay fallback to open a direct, authenticated QUIC connection between any two nodes regardless of network location. There is no central server; Mesh LLM adds its own gossip layer on top of iroh's transport to control mesh admission, version compatibility, and peer trust. Developers evaluating peer-to-peer transport layers for distributed apps can track approaches like this on daily.dev."}}]} ```