<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/ollama-integrates-with-claude-desktop-to-run-open-weight-models-locally-mwe4re20x" -->

---
title: Ollama integrates with Claude Desktop to run open-weight...
description: Ollama v0.33.0 now integrates with Claude Desktop, letting users route requests to open-weight models like Qwen, DeepSeek, Kimi, and GLM directly from within...
canonical: https://daily.dev/posts/ollama-integrates-with-claude-desktop-to-run-open-weight-models-locally-mwe4re20x
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: Ollama integrates with Claude Desktop to run open-weight models locally | daily.dev
og:description: Ollama v0.33.0 now integrates with Claude Desktop, letting users route requests to open-weight models like Qwen, DeepSeek, Kimi, and GLM directly from within...
og:url: https://daily.dev/posts/ollama-integrates-with-claude-desktop-to-run-open-weight-models-locally-mwe4re20x
og:image: https://api.daily.dev/og/posts/MWE4rE20x.png
og:image:alt: Ollama integrates with Claude Desktop to run open-weight models locally
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.

# Ollama integrates with Claude Desktop to run open-weight models locally

**[Collections](https://daily.dev/sources/collections)** · 1 min read · 0 upvotes · 0 comments

## Summary

Ollama v0.33.0 now integrates with Claude Desktop, letting users route requests to open-weight models like Qwen, DeepSeek, Kimi, and GLM directly from within Anthropic's app, running locally or via Ollama Cloud. Setup happens through Ollama's Mac app, mapping models like Kimi K3 or DeepSeek V4 Pro into Claude's Opus and Sonnet slots. This is a second attempt after an earlier April integration broke when Claude Desktop began rejecting non-Anthropic model IDs; the new version works around that with a dedicated local proxy. Telemetry is off by default with zero data retention. The feature is Mac-only for now, with Windows support possibly coming later. The release follows a $65 million Series B funding round for Ollama.

## Content

Ollama has added Claude Desktop support in v0.33.0, letting users route requests to open-weight models like Qwen, DeepSeek, Kimi, and GLM directly from Anthropic's app. Models can run locally or through Ollama Cloud, and switching back to Anthropic's own models is straightforward.

This is actually Ollama's second attempt at this. An earlier integration in April used a third-party gateway setting, but broke within days when Claude Desktop started rejecting non-Anthropic model IDs. The new release works around that by running a dedicated local proxy instead.

Setup is handled through Ollama's Mac app. Users can map open-weight models to Claude's Opus and Sonnet slots - so something like Kimi K3 or DeepSeek V4 Pro can sit where Claude Opus normally would. Telemetry is off by default, and Ollama enforces a zero data retention policy. The idea is you use cloud models for heavier tasks and local models for everything else, swapping between them as needed.

The feature is currently Mac-only. Windows support may come later. The release follows Ollama's $65 million Series B funding round.

## Questions this post answers

### How does Ollama v0.33.0 integrate open-weight models into Claude Desktop?

Ollama v0.33.0 adds a dedicated local proxy that lets Claude Desktop route requests to open-weight models like Qwen, DeepSeek, Kimi, and GLM, running them locally or via Ollama Cloud. Users map these models to Claude's Opus and Sonnet slots through Ollama's Mac app, and can switch back to Anthropic's own models at any time. The feature is Mac-only, with Windows support possibly coming later.

_Developers mixing local and cloud models in Claude Desktop can follow tool integrations like this on daily.dev._

### Why did Ollama's earlier Claude Desktop integration stop working?

An earlier integration released in April used a third-party gateway setting to route non-Anthropic models into Claude Desktop, but it broke within days after Claude Desktop started rejecting non-Anthropic model IDs. The v0.33.0 release fixes this by running a dedicated local proxy instead of relying on a gateway setting that Anthropic's app could block.

_Anyone relying on third-party integrations with closed apps can track breaking changes like this via daily.dev._

### Does Ollama's Claude Desktop integration send my data anywhere?

Telemetry is off by default and Ollama enforces a zero data retention policy for this integration, whether models run locally or through Ollama Cloud. This lets users pick heavier tasks for cloud-hosted open-weight models while keeping everything else on local models, without persistent data logging.

_Developers weighing privacy trade-offs between local and cloud inference can follow updates like this on daily.dev._

## Similar posts on daily.dev

- [Untitled](https://daily.dev/posts/untitled-sn7w6mzum) · SitePoint · 0 upvotes · 0 comments

---

Tags: [#open-source](https://daily.dev/tags/open-source), [#local-ai](https://daily.dev/tags/local-ai), [#ollama](https://daily.dev/tags/ollama)

[View this post on daily.dev](https://daily.dev/posts/ollama-integrates-with-claude-desktop-to-run-open-weight-models-locally-mwe4re20x)

```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":"Ollama integrates with Claude Desktop to run open-weight models locally","url":"https://daily.dev/posts/ollama-integrates-with-claude-desktop-to-run-open-weight-models-locally-mwe4re20x","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/ollama-integrates-with-claude-desktop-to-run-open-weight-models-locally-mwe4re20x"},"datePublished":"2026-08-26T17:08:15.238Z","dateModified":"2026-10-15T02:22:38.723Z","description":"Ollama v0.33.0 now integrates with Claude Desktop, letting users route requests to open-weight models like Qwen, DeepSeek, Kimi, and GLM directly from within...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/71151f216f12349306e2e60bff1e3aa1?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/71151f216f12349306e2e60bff1e3aa1?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Collections","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":"Collections","logo":"https://media.daily.dev/image/upload/s--fk_6ycEi--/f_auto,q_auto/v1780996001/logos/collections?_a=BAMAMiWQ0","url":"https://daily.dev/sources/collections"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/ollama-integrates-with-claude-desktop-to-run-open-weight-models-locally-mwe4re20x","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"open-source,local-ai,ollama","timeRequired":"PT1M"}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Collections","item":"https://daily.dev/sources/collections"},{"@type":"ListItem","position":3,"name":"Ollama integrates with Claude Desktop to run open-weight models locally"}]}
{"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/ollama-integrates-with-claude-desktop-to-run-open-weight-models-locally-mwe4re20x#faq","mainEntity":[{"@type":"Question","name":"How does Ollama v0.33.0 integrate open-weight models into Claude Desktop?","acceptedAnswer":{"@type":"Answer","text":"Ollama v0.33.0 adds a dedicated local proxy that lets Claude Desktop route requests to open-weight models like Qwen, DeepSeek, Kimi, and GLM, running them locally or via Ollama Cloud. Users map these models to Claude's Opus and Sonnet slots through Ollama's Mac app, and can switch back to Anthropic's own models at any time. The feature is Mac-only, with Windows support possibly coming later. Developers mixing local and cloud models in Claude Desktop can follow tool integrations like this on daily.dev."}},{"@type":"Question","name":"Why did Ollama's earlier Claude Desktop integration stop working?","acceptedAnswer":{"@type":"Answer","text":"An earlier integration released in April used a third-party gateway setting to route non-Anthropic models into Claude Desktop, but it broke within days after Claude Desktop started rejecting non-Anthropic model IDs. The v0.33.0 release fixes this by running a dedicated local proxy instead of relying on a gateway setting that Anthropic's app could block. Anyone relying on third-party integrations with closed apps can track breaking changes like this via daily.dev."}},{"@type":"Question","name":"Does Ollama's Claude Desktop integration send my data anywhere?","acceptedAnswer":{"@type":"Answer","text":"Telemetry is off by default and Ollama enforces a zero data retention policy for this integration, whether models run locally or through Ollama Cloud. This lets users pick heavier tasks for cloud-hosted open-weight models while keeping everything else on local models, without persistent data logging. Developers weighing privacy trade-offs between local and cloud inference can follow updates like this on daily.dev."}}]}
```

