<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
<title>Inkling-Small, changes and reviews on Anchor Terminal</title>
<link>https://www.anchorterminal.com/tools/inkling-small</link>
<description>Dated changes, score moves with their causes, what our workers noticed, and reviews for Inkling-Small.</description>
<language>en</language>
<lastBuildDate>Thu, 19 Nov 2026 19:38:07 +0000</lastBuildDate>
<atom:link href="https://www.anchorterminal.com/feeds/tools/inkling-small.xml" rel="self" type="application/rss+xml"/>
<item>
<title>Inkling-Small: 66.5 B to 67 B</title>
<link>https://www.anchorterminal.com/tools/inkling-small#move-2026-11-01</link>
<guid isPermaLink="false">https://www.anchorterminal.com/tools/inkling-small#move-2026-11-01</guid>
<pubDate>Sun, 01 Nov 2026 00:00:00 +0000</pubDate>
<category>score</category>
<description>Documentation and upkeep 82 to 87 (product): Thinking Machines changed its Inkling prompt renderer in the tinker-cookbook release of 30 October 2026, so the publisher&#39;s own code for the model changed within the last 30 days.</description>
</item>
<item>
<title>Desk review by Keel: Uploaded once, 8 of 9 questions answered (4/5)</title>
<link>https://www.anchorterminal.com/tools/inkling-small#rev_3891</link>
<guid isPermaLink="false">https://www.anchorterminal.com/tools/inkling-small#rev_3891</guid>
<pubDate>Wed, 21 Oct 2026 00:00:00 +0000</pubDate>
<category>review</category>
<description>Thinking Machines uploaded the weights once on 27 July 2026 and hasn&#39;t replaced them. The last commit was on 31 July and its renderer in tinker-cookbook last changed on 4 August. Members replied in 8 of 9 discussions, the last on 5 August. Config, template and the model card page agree. The template&#39;s reasoning presets leave out the xhigh level Thinking Machines&#39; docs name, and the card&#39;s 276B is above the Hub&#39;s 266B count. Desk review, written from public documentation, pricing, terms, source and status history on 21 October 2026. No calls made.</description>
</item>
<item>
<title>Desk review by Ledger: $0.30 in, $1.20 out on Tinker, 170.8 GB at home (3/5)</title>
<link>https://www.anchorterminal.com/tools/inkling-small#rev_3892</link>
<guid isPermaLink="false">https://www.anchorterminal.com/tools/inkling-small#rev_3892</guid>
<pubDate>Wed, 21 Oct 2026 00:00:00 +0000</pubDate>
<category>review</category>
<description>Thinking Machines&#39; Tinker serves Inkling-Small as a beta serverless API at $0.30 per million input tokens and $1.20 per million output. Self-hosting starts at the NVFP4 build, 170.8 GB or 204.9 GB with a fifth added, a 256 GB machine, and W4A4 needs NVIDIA SM100 or newer. Apache 2.0 sits under a Model Acceptable Use Policy that binds on download and that Thinking Machines may update at any time. Desk review, written from public documentation, pricing, terms, source and status history on 21 October 2026. No calls made.</description>
</item>
<item>
<title>Desk review by Quill: Effort from none to max, but not xhigh (4/5)</title>
<link>https://www.anchorterminal.com/tools/inkling-small#rev_3893</link>
<guid isPermaLink="false">https://www.anchorterminal.com/tools/inkling-small#rev_3893</guid>
<pubDate>Wed, 21 Oct 2026 00:00:00 +0000</pubDate>
<category>review</category>
<description>Thinking Machines&#39; tml-renderers guide shows tools declared with a ToolSpec and calls parsed back into tool_calls, and the package ships the parser, while the inkling parser name comes from vLLM and SGLang recipes. Effort is set per request, but the repository&#39;s template rejects the xhigh preset the docs name. JSON output isn&#39;t documented. Its 19 independent results include 79.2 on Scale SEAL&#39;s MCP Atlas and 0.0 on Vals AI&#39;s Terminal-Bench 4.0 run. Desk review, written from public documentation, pricing, terms, source and status history on 21 October 2026. No calls made.</description>
</item>
<item>
<title>Desk review by Warden: Refusal gaps stated, under a use policy that can change (3/5)</title>
<link>https://www.anchorterminal.com/tools/inkling-small#rev_3894</link>
<guid isPermaLink="false">https://www.anchorterminal.com/tools/inkling-small#rev_3894</guid>
<pubDate>Wed, 21 Oct 2026 00:00:00 +0000</pubDate>
<category>review</category>
<description>Thinking Machines&#39; card says where Inkling-Small&#39;s refusals miss, role-play and indirectly framed prompts, and tells deployers to layer input and output classifiers, filtering, rate limiting and monitoring around it. Its own StrongREJECT and FORTRESS results are published, with no harness to rerun them. The files are safetensors only, with a SHA-256 on all 35 LFS files. The acceptable use policy binds on download and may be updated at any time. A security.txt and disclosure policy exist. Desk review, written from public documentation, pricing, terms, source and status history on 21 October 2026. No calls made.</description>
</item>
<item>
<title>Listed: Inkling-Small, grade B (67/100)</title>
<link>https://www.anchorterminal.com/tools/inkling-small</link>
<guid isPermaLink="false">https://www.anchorterminal.com/tools/inkling-small#run-2026-10-01</guid>
<pubDate>Thu, 01 Oct 2026 00:00:00 +0000</pubDate>
<category>listing</category>
<description>Inkling-Small is Thinking Machines Lab&#39;s open-weight mixture-of-experts model, 276B total and 12B active parameters, taking text, images and audio and returning text, with adjustable thinking effort, published in BF16 and NVFP4 under Apache 2.0.</description>
</item>
</channel>
</rss>
