<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Silk Compute blog</title>
    <link>https://silkcompute.com/blog</link>
    <atom:link href="https://silkcompute.com/rss.xml" rel="self" type="application/rss+xml"/>
    <description>How abliterated models work, how we compare, and what's new.</description>
    <language>en-us</language>
    <item>
      <title>Best Uncensored Models for Coding Agents: How to Pick One</title>
      <link>https://silkcompute.com/blog/uncensored-model-for-coding-agents</link>
      <guid>https://silkcompute.com/blog/uncensored-model-for-coding-agents</guid>
      <pubDate>Tue, 10 Nov 2026 00:00:00 GMT</pubDate>
      <category>Guide</category>
      <description>The best abliterated models for Claude Code, Codex and OpenCode: what an agent needs, which model suits each job, and what each one really costs.</description>
    </item>
    <item>
      <title>Kimi K3.1 Pro Uncensored: Specs, Pricing and Best Uses</title>
      <link>https://silkcompute.com/blog/kimi-k3-1-uncensored</link>
      <guid>https://silkcompute.com/blog/kimi-k3-1-uncensored</guid>
      <pubDate>Tue, 10 Nov 2026 00:00:00 GMT</pubDate>
      <category>Launch</category>
      <description>Abliterated Kimi K3.1 Pro with image and video input, 1M tokens of context and up to 131K output tokens: specs, pricing, best uses and setup.</description>
    </item>
    <item>
      <title>Uncensored LLM for Synthetic Data: A Guide to the Datasets Other Models Refuse</title>
      <link>https://silkcompute.com/blog/uncensored-llm-synthetic-data</link>
      <guid>https://silkcompute.com/blog/uncensored-llm-synthetic-data</guid>
      <pubDate>Mon, 09 Nov 2026 00:00:00 GMT</pubDate>
      <category>Guide</category>
      <description>Generate the synthetic data filtered models refuse with an uncensored LLM: cost per 1,000 rows on five models, picks by dataset and a clean pipeline.</description>
    </item>
    <item>
      <title>Abliterated AI: A Complete Guide to Choosing a Model</title>
      <link>https://silkcompute.com/blog/what-is-an-abliterated-model</link>
      <guid>https://silkcompute.com/blog/what-is-an-abliterated-model</guid>
      <pubDate>Thu, 05 Nov 2026 00:00:00 GMT</pubDate>
      <category>Guide</category>
      <description>Explore abliterated AI models, from laptop to frontier. Use our size chart to pick the best uncensored model for coding, security work or running at home.</description>
    </item>
    <item>
      <title>Hy4 Uncensored: Specs, Pricing and Best Uses</title>
      <link>https://silkcompute.com/blog/hy4-uncensored</link>
      <guid>https://silkcompute.com/blog/hy4-uncensored</guid>
      <pubDate>Thu, 05 Nov 2026 00:00:00 GMT</pubDate>
      <category>Launch</category>
      <description>Abliterated Tencent Hy4 with image input, 1M tokens of context and up to 66K output tokens: specs, pricing, benchmarks, best uses and setup.</description>
    </item>
    <item>
      <title>Venice AI API alternative for uncensored models</title>
      <link>https://silkcompute.com/blog/venice-ai-api-alternative</link>
      <guid>https://silkcompute.com/blog/venice-ai-api-alternative</guid>
      <pubDate>Wed, 04 Nov 2026 00:00:00 GMT</pubDate>
      <category>Comparison</category>
      <description>Need a Venice AI API alternative? Run abliterated DeepSeek and GLM in Claude Code with no router, on plans from $10 a month. See how the two compare.</description>
    </item>
    <item>
      <title>abliteration.ai alternative: DeepSeek and GLM for less</title>
      <link>https://silkcompute.com/blog/abliteration-ai-alternative</link>
      <guid>https://silkcompute.com/blog/abliteration-ai-alternative</guid>
      <pubDate>Wed, 04 Nov 2026 00:00:00 GMT</pubDate>
      <category>Comparison</category>
      <description>Need an abliteration.ai alternative? Get the same DeepSeek and GLM-5.3 base models at lower input and output prices, on plans from $10 a month.</description>
    </item>
    <item>
      <title>Does OpenRouter have uncensored or abliterated models?</title>
      <link>https://silkcompute.com/blog/openrouter-uncensored-models</link>
      <guid>https://silkcompute.com/blog/openrouter-uncensored-models</guid>
      <pubDate>Wed, 04 Nov 2026 00:00:00 GMT</pubDate>
      <category>Comparison</category>
      <description>As of November 2026, OpenRouter's public list calls none of its models abliterated and two 24B fine-tunes uncensored. Why routing doesn't remove refusals.</description>
    </item>
    <item>
      <title>DeepSeek V4.1 Flash 1017 Uncensored is live</title>
      <link>https://silkcompute.com/blog/deepseek-v4-1-flash-uncensored</link>
      <guid>https://silkcompute.com/blog/deepseek-v4-1-flash-uncensored</guid>
      <pubDate>Wed, 04 Nov 2026 00:00:00 GMT</pubDate>
      <category>Launch</category>
      <description>DeepSeek V4.1 Flash 1017 Uncensored is live: 1M tokens of context, low, high or max effort, and the lowest agent bill of our models.</description>
    </item>
    <item>
      <title>GLM-5.3 Flash Uncensored is live</title>
      <link>https://silkcompute.com/blog/glm-5-3-flash-uncensored</link>
      <guid>https://silkcompute.com/blog/glm-5-3-flash-uncensored</guid>
      <pubDate>Wed, 04 Nov 2026 00:00:00 GMT</pubDate>
      <category>Launch</category>
      <description>GLM-5.3 Flash Uncensored is live: the lowest input price of our models, with 1M tokens of context, made for batch jobs and background work.</description>
    </item>
  </channel>
</rss>
