---
title: "Thinking Machines releases Inkling 2 Mini, an open model that runs on a laptop | Coding Delta"
description: "Thinking Machines Lab published Inkling 2 Mini, a 26.4B-parameter mixture-of-experts model with open weights under Apache 2.0, aimed at coding and agent work with a memory footprint of about 20GB. It is already available through Ollama, vLLM, SGLang and quantised GGUF builds."
canonical_url: "https://codingdelta.com/news/stories/32-thinking-machines-releases-inkling-2-mini-an-open-model-that-runs-on-a"
last_updated: "2026-11-21T12:47:23+01:00"
locale: en
meta:
  "og:description": "Thinking Machines Lab published Inkling 2 Mini, a 26.4B-parameter mixture-of-experts model with open weights under Apache 2.0, aimed at coding and agent work with a memory footprint of about 20GB. It is already available through Ollama, vLLM, SGLang and quantised GGUF builds."
  "og:image": "https://codingdelta.com/og-image.png"
  "og:image:alt": "Turn your ideas into reality with Coding Delta, your boutique software development agency"
  "og:image:height": "630"
  "og:image:type": image/png
  "og:image:width": "1200"
  "og:title": "Thinking Machines releases Inkling 2 Mini, an open model that runs on a laptop"
  "twitter:image": "https://codingdelta.com/og-image.png"
  "twitter:image:alt": "Turn your ideas into reality with Coding Delta, your boutique software development agency"
---

> Index of every page on this site: [llms.txt](https://codingdelta.com/llms.txt)

[AI & automation · 21 November 2026](https://codingdelta.com/news)

# **Thinking Machines releases Inkling 2 Mini, an open model that runs on a laptop**

[How this briefing is made](https://codingdelta.com/news/about)

[Covered in edition 21 November 2026](https://codingdelta.com/news/editions/2026-11-21#item-32)

Thinking Machines Lab published Inkling 2 Mini, a 26.4B-parameter mixture-of-experts model with open weights under Apache 2.0, aimed at coding and agent work with a memory footprint of about 20GB. It is already available through Ollama, vLLM, SGLang and quantised GGUF builds.

**Why it matters:** Teams that cannot send code or customer data to a US cloud API now have another capable coding model they can run on their own hardware; worth a test against your current assistant.

[Source: Thinking Machines Lab](https://thinkingmachines.ai/news/inkling-2-mini/) [#Open source](https://codingdelta.com/news/topics/open-source) [#AI models](https://codingdelta.com/news/topics/ai-models)

## Also covered by

- [**X (@thinkymachines)**](https://x.com/thinkymachines/status/2123871643797995872)
- [**X (@ollama)**](https://x.com/ollama/status/2123889582303281557)
- [**Hugging Face**: thinkingmachines/Inkling-2-Mini](https://huggingface.co/thinkingmachines/Inkling-2-Mini)

## **How the story developed**

1. 29 October 2026 [Google announces Gemini 4 Carbon, its new frontier model](https://codingdelta.com/news/stories/20-google-announces-gemini-4-carbon-its-new-frontier-model)
2. 13 November 2026 [Google releases Gemini 4 Hydrogen for fast, low-cost coding and agent work](https://codingdelta.com/news/stories/24-google-releases-gemini-4-hydrogen-for-fast-low-cost-coding-and-agent-w)
3. 20 November 2026 [Google launches Gemini 4 Neon, a cheaper mid-tier model for coding and agents](https://codingdelta.com/news/stories/18-google-launches-gemini-4-neon-a-cheaper-mid-tier-model-for-coding-and)
4. 21 November 2026

   **Thinking Machines releases Inkling 2 Mini, an open model that runs on a laptop**

```json
{"@context":"https://schema.org","@graph":[{"@id":"https://codingdelta.com/#website","@type":"WebSite","description":"Boutique software studio in Amsterdam for web apps, platforms, mobile apps and AI agents.","inLanguage":"en-US","name":"Coding Delta","url":"https://codingdelta.com/","publisher":{"@id":"https://codingdelta.com/#identity"},"workTranslation":{"@id":"https://codingdelta.com/nl#website"}},{"@id":"https://codingdelta.com/news/stories/32-thinking-machines-releases-inkling-2-mini-an-open-model-that-runs-on-a#webpage","@type":"WebPage","dateModified":"2026-11-21T12:47:23+01:00","datePublished":"2026-11-21T12:47:23+01:00","description":"Thinking Machines Lab published Inkling 2 Mini, a 26.4B-parameter mixture-of-experts model with open weights under Apache 2.0, aimed at coding and agent work with a memory footprint of about 20GB. It is already available through Ollama, vLLM, SGLang and quantised GGUF builds.","name":"Thinking Machines releases Inkling 2 Mini, an open model that runs on a laptop","url":"https://codingdelta.com/news/stories/32-thinking-machines-releases-inkling-2-mini-an-open-model-that-runs-on-a","breadcrumb":{"@id":"https://codingdelta.com/news/stories/32-thinking-machines-releases-inkling-2-mini-an-open-model-that-runs-on-a#breadcrumb"},"isPartOf":{"@id":"https://codingdelta.com/#website"},"potentialAction":[{"@type":"ReadAction","target":["https://codingdelta.com/news/stories/32-thinking-machines-releases-inkling-2-mini-an-open-model-that-runs-on-a"]}],"primaryImageOfPage":{"@id":"https://codingdelta.com/#logo"}},{"@id":"https://codingdelta.com/#identity","description":"Boutique software studio in Amsterdam: strategy and scoping, product design, engineering (web apps, platforms, mobile) and AI & automation.","iso6523Code":"0106:77615999","legalName":"Coding Delta B.V.","name":"Coding Delta","url":"https://codingdelta.com","@type":["Organization","LocalBusiness","ProfessionalService"],"address":{"@type":"PostalAddress","streetAddress":"Keizersgracht 241","postalCode":"1016 EA","addressLocality":"Amsterdam","addressCountry":"NL"},"brand":[{"@type":"Brand","name":"Chatsome","url":"https://chatsome.co/","description":"AI agents for customer service and sales"},{"@type":"Brand","name":"Delta Security","url":"https://delta.codingdelta.com/","description":"Website security scans"},{"@type":"Brand","name":"RoundE","url":"https://rounde.nl/","description":"Employee participation for startups"}],"identifier":{"@type":"PropertyValue","propertyID":"KVK","value":"77615999"},"image":{"@id":"https://codingdelta.com/#/schema/image/1"},"knowsAbout":["Software development","Web application development","Product design","Mobile app development","AI agents","Business process automation"],"knowsLanguage":["en","nl"],"sameAs":["https://www.linkedin.com/company/codingdelta/"]},{"@id":"https://codingdelta.com/news/stories/32-thinking-machines-releases-inkling-2-mini-an-open-model-that-runs-on-a#breadcrumb","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","name":"Home","item":"https://codingdelta.com","position":1},{"@type":"ListItem","name":"Coding Delta News","item":"https://codingdelta.com/news","position":2},{"@type":"ListItem","name":"Thinking Machines releases Inkling 2 Mini, an open model that runs on a laptop","position":3}]},{"@id":"https://codingdelta.com/news/stories/32-thinking-machines-releases-inkling-2-mini-an-open-model-that-runs-on-a#article","dateModified":"2026-11-21T12:47:23+01:00","datePublished":"2026-11-21T12:47:23+01:00","description":"Thinking Machines Lab published Inkling 2 Mini, a 26.4B-parameter mixture-of-experts model with open weights under Apache 2.0, aimed at coding and agent work with a memory footprint of about 20GB. It is already available through Ollama, vLLM, SGLang and quantised GGUF builds.","headline":"Thinking Machines releases Inkling 2 Mini, an open model that runs on a laptop","inLanguage":"en-US","isBasedOn":"https://thinkingmachines.ai/news/inkling-2-mini/?utm_source=news.codingdelta.com&utm_medium=web","thumbnailUrl":"og-image.png","@type":["Article","NewsArticle"],"articleSection":["AI & automation"],"author":{"@id":"https://codingdelta.com/#identity"},"image":{"@id":"https://codingdelta.com/#/schema/image/2"},"isPartOf":{"@id":"https://codingdelta.com/news/stories/32-thinking-machines-releases-inkling-2-mini-an-open-model-that-runs-on-a#webpage"},"keywords":["Open source","AI models"],"mainEntityOfPage":{"@id":"https://codingdelta.com/news/stories/32-thinking-machines-releases-inkling-2-mini-an-open-model-that-runs-on-a#webpage"},"publisher":{"@id":"https://codingdelta.com/#identity"}},{"@id":"https://codingdelta.com/#/schema/image/1","@type":"ImageObject","contentUrl":"https://codingdelta.com/og-image.png","inLanguage":"en-US","url":"https://codingdelta.com/og-image.png"},{"@id":"https://codingdelta.com/#logo","@type":"ImageObject","caption":"Coding Delta","contentUrl":"https://codingdelta.com/logo.png","inLanguage":"en-US","url":"https://codingdelta.com/logo.png"},{"@id":"https://codingdelta.com/#organization","@type":"Organization","logo":"https://codingdelta.com/logo.png","name":"Coding Delta","url":"https://codingdelta.com","address":{"@type":"PostalAddress","streetAddress":"Keizersgracht 241","postalCode":"1016 EA","addressLocality":"Amsterdam","addressCountry":"NL"},"sameAs":["https://www.linkedin.com/company/codingdelta/"]},{"@id":"https://codingdelta.com/#/schema/image/2","@type":"ImageObject","contentUrl":"og-image.png","inLanguage":"en-US","url":"og-image.png"}]}
```

## Sitemap

See the full [sitemap](https://codingdelta.com/sitemap.md) for all pages.
