<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
  <title>MATRIX · Pika H3 Flex</title>
  <link>https://derekwang2002.github.io/ai-intelligence/en/projects/pika-h3-flex/</link>
  <description>Material changes in AI models, agents, engineering and infrastructure, with primary sources and recommendations.</description>
  <language>en</language>
  <lastBuildDate>Mon, 02 Nov 2026 16:56:00 GMT</lastBuildDate>
  <atom:link href="https://derekwang2002.github.io/ai-intelligence/en/feeds/projects/pika-h3-flex.xml" rel="self" type="application/rss+xml"/>
  <item>
    <title>Pika Labs releases Pika H3 Flex video-model weights</title>
    <link>https://derekwang2002.github.io/ai-intelligence/en/events/ev-20261029-05/</link>
    <guid isPermaLink="false">ev-20261029-05-published</guid>
    <pubDate>Fri, 30 Oct 2026 16:01:56 GMT</pubDate>
    <category>New</category>
    <category>TRIAL</category>
    <description>&lt;p&gt;Pika Labs released ungated Pika H3 Flex weights as a LoRA over MiniMax H3 for local image-to-video generation with synchronized audio. The model card advertises six-step generation, up to 12 references and native 480p/768p output; hosted API features extend to higher resolutions. The repository uses a custom MiniMax H3 community license rather than Apache or MIT.&lt;/p&gt;&lt;p&gt;Recommendation: TRIAL&lt;/p&gt;&lt;p&gt;Sources: &lt;a href=&quot;https://huggingface.co/Pika-Labs/Pika-H3-Flex&quot;&gt;Pika H3 Flex&lt;/a&gt; · &lt;a href=&quot;https://github.com/pika-labs/Pika-H3-Flex&quot;&gt;Pika H3 Flex code&lt;/a&gt;&lt;/p&gt;</description>
  </item>
</channel>
</rss>
