<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Zabio news: model, price and vendor changes</title>
    <link>https://zabio.ai/news/</link>
    <description>What the model vendors announced, what it means for your calls, and what Zabio did about it.</description>
    <language>en</language>
    <atom:link href="https://zabio.ai/news/feed.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>grok-build-0.1 reached end of life on November 19, 2026</title>
      <link>https://zabio.ai/news/2026-11-20-xai-grok-build-0-1-reached-end-of-life-on-november-19-2026/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-20-xai-grok-build-0-1-reached-end-of-life-on-november-19-2026/</guid>
      <pubDate>Fri, 20 Nov 2026 19:10:16 GMT</pubDate>
      <category>xai</category>
      <description>xAI retired grok-build-0.1 on November 19, 2026, and the API no longer serves it.</description>
    </item>
    <item>
      <title>Qwen3 TTS now available on Together AI</title>
      <link>https://zabio.ai/news/2026-11-20-together-qwen3-tts-now-available-on-together-ai/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-20-together-qwen3-tts-now-available-on-together-ai/</guid>
      <pubDate>Fri, 20 Nov 2026 16:55:22 GMT</pubDate>
      <category>together</category>
      <description>Qwen's open text-to-speech model is now on Together AI, on serverless inference and dedicated endpoints, at $20 per 1M characters.</description>
    </item>
    <item>
      <title>Together AI adds tencent/Hy4 serverless model with 1,048,576 context</title>
      <link>https://zabio.ai/news/2026-11-19-together-together-ai-adds-tencent-hy4-serverless-model-with-1-048/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-19-together-together-ai-adds-tencent-hy4-serverless-model-with-1-048/</guid>
      <pubDate>Thu, 19 Nov 2026 22:50:17 GMT</pubDate>
      <category>together</category>
      <description>Together AI added tencent/Hy4 to its serverless models, with a 1,048,576 context length and per-token pricing.</description>
    </item>
    <item>
      <title>Tool Retrieval in the Gemini API (preview)</title>
      <link>https://zabio.ai/news/2026-11-19-google-tool-retrieval-in-the-gemini-api-preview/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-19-google-tool-retrieval-in-the-gemini-api-preview/</guid>
      <pubDate>Thu, 19 Nov 2026 18:40:21 GMT</pubDate>
      <category>google</category>
      <description>Google adds Tool Retrieval to the Gemini API in preview, loading tool schemas on demand to reduce input tokens.</description>
    </item>
    <item>
      <title>Together AI adds Qwen/Qwen4-Max as a serverless model</title>
      <link>https://zabio.ai/news/2026-11-18-together-together-ai-adds-qwen-qwen4-max-as-a-serverless-model/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-18-together-together-ai-adds-qwen-qwen4-max-as-a-serverless-model/</guid>
      <pubDate>Wed, 18 Nov 2026 06:15:18 GMT</pubDate>
      <category>together</category>
      <description>Together AI added Qwen/Qwen4-Max, with a 1,000,000 context length and text and image input.</description>
    </item>
    <item>
      <title>Two NVIDIA Nemotron 3.5 models now available for fine-tuning</title>
      <link>https://zabio.ai/news/2026-11-17-together-two-nvidia-nemotron-3-5-models-now-available-for-fine-tuning/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-17-together-two-nvidia-nemotron-3-5-models-now-available-for-fine-tuning/</guid>
      <pubDate>Tue, 17 Nov 2026 19:25:19 GMT</pubDate>
      <category>together</category>
      <description>Together AI now lists two NVIDIA Nemotron 3.5 models for fine-tuning, with support for function-calling and reasoning.</description>
    </item>
    <item>
      <title>2h and 8h background windows in Responses API beta: 30% and 40% lower token rates</title>
      <link>https://zabio.ai/news/2026-11-17-openai-2h-and-8h-background-windows-in-responses-api-beta-30-and-40/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-17-openai-2h-and-8h-background-windows-in-responses-api-beta-30-and-40/</guid>
      <pubDate>Tue, 17 Nov 2026 17:50:48 GMT</pubDate>
      <category>openai</category>
      <description>OpenAI adds 2h and 8h completion windows for background Responses API requests, with model token rates 30% and 40% lower than Standard.</description>
    </item>
    <item>
      <title>OpenAI adds GPT-6.2 Astra and GPT-6.2 Sol to its API</title>
      <link>https://zabio.ai/news/2026-11-17-openai-openai-adds-gpt-6-2-astra-and-gpt-6-2-sol-to-its-api/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-17-openai-openai-adds-gpt-6-2-astra-and-gpt-6-2-sol-to-its-api/</guid>
      <pubDate>Tue, 17 Nov 2026 17:50:26 GMT</pubDate>
      <category>openai</category>
      <description>OpenAI's API now offers GPT-6.2 Astra and GPT-6.2 Sol, both with a 1,050,000 token context window.</description>
    </item>
    <item>
      <title>Gemini 4 Hydrogen: a fast robotics model now in the Gemini API</title>
      <link>https://zabio.ai/news/2026-11-17-google-gemini-4-hydrogen-a-fast-robotics-model-now-in-the-gemini-api/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-17-google-gemini-4-hydrogen-a-fast-robotics-model-now-in-the-gemini-api/</guid>
      <pubDate>Tue, 17 Nov 2026 17:45:31 GMT</pubDate>
      <category>google</category>
      <description>Google released Gemini 4 Hydrogen, a fast model for robotics, with the model ID gemini-4-hydrogen.</description>
    </item>
    <item>
      <title>DeepSeek opens app features like web search and file input in its API</title>
      <link>https://zabio.ai/news/2026-11-13-deepseek-deepseek-opens-app-features-like-web-search-and-file-input/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-13-deepseek-deepseek-opens-app-features-like-web-search-and-file-input/</guid>
      <pubDate>Fri, 13 Nov 2026 14:20:31 GMT</pubDate>
      <category>deepseek</category>
      <description>deepseek-flash now supports web search, deep thinking search, file input, URL reading and a safety check in three APIs.</description>
    </item>
    <item>
      <title>Together AI adds thinkingmachines/Inkling-2-Small</title>
      <link>https://zabio.ai/news/2026-11-12-together-together-ai-adds-thinkingmachines-inkling-2-small/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-12-together-together-ai-adds-thinkingmachines-inkling-2-small/</guid>
      <pubDate>Thu, 12 Nov 2026 20:50:18 GMT</pubDate>
      <category>together</category>
      <description>Together AI added the serverless model thinkingmachines/Inkling-2-Small, with a 1,048,576 context length and text and image input.</description>
    </item>
    <item>
      <title>New serverless model elorian-labs/glimpse-1.0 on Together AI</title>
      <link>https://zabio.ai/news/2026-11-12-together-new-serverless-model-elorian-labs-glimpse-1-0-on-together-ai/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-12-together-new-serverless-model-elorian-labs-glimpse-1-0-on-together-ai/</guid>
      <pubDate>Thu, 12 Nov 2026 16:10:18 GMT</pubDate>
      <category>together</category>
      <description>Together AI added the serverless model elorian-labs/glimpse-1.0, with a 262,144 context length and listed input, output and cached input prices.</description>
    </item>
    <item>
      <title>Gemini Decision 1 model and Decisions API for fast, calibrated decisions</title>
      <link>https://zabio.ai/news/2026-11-11-google-gemini-decision-1-model-and-decisions-api-for-fast-calibrated/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-11-google-gemini-decision-1-model-and-decisions-api-for-fast-calibrated/</guid>
      <pubDate>Wed, 11 Nov 2026 18:20:29 GMT</pubDate>
      <category>google</category>
      <description>Google releases gemini-decision-1 in the Gemini API for yes/no, multiple choice and rubric questions at $0.10 per 1M input tokens.</description>
    </item>
    <item>
      <title>Claude Fable 5.2 and Claude Mythos 5.2 read other models' thinking blocks</title>
      <link>https://zabio.ai/news/2026-11-11-anthropic-claude-fable-5-2-and-claude-mythos-5-2-read-other-models/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-11-anthropic-claude-fable-5-2-and-claude-mythos-5-2-read-other-models/</guid>
      <pubDate>Wed, 11 Nov 2026 17:15:41 GMT</pubDate>
      <category>anthropic</category>
      <description>Claude Fable 5.2 and Claude Mythos 5.2 read thinking blocks from Claude Opus 5.5, Claude Sonnet 5.5, and Claude Haiku 5.5 on the API.</description>
    </item>
    <item>
      <title>Claude Fable 5.2 and Claude Mythos 5.2 use a new computer use toolset</title>
      <link>https://zabio.ai/news/2026-11-09-anthropic-claude-fable-5-2-and-claude-mythos-5-2-use-a-new-computer/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-09-anthropic-claude-fable-5-2-and-claude-mythos-5-2-use-a-new-computer/</guid>
      <pubDate>Mon, 09 Nov 2026 19:35:21 GMT</pubDate>
      <category>anthropic</category>
      <description>Claude Fable 5.2 and Claude Mythos 5.2 support computer use through a new toolset, and the earlier computer tool returns a 400 error on most platforms.</description>
    </item>
    <item>
      <title>Gemini Omni 1.2 Flash: 15-second clips and 30% lower price</title>
      <link>https://zabio.ai/news/2026-11-05-google-gemini-omni-1-2-flash-15-second-clips-and-30-lower-price/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-05-google-gemini-omni-1-2-flash-15-second-clips-and-30-lower-price/</guid>
      <pubDate>Thu, 05 Nov 2026 17:10:45 GMT</pubDate>
      <category>google</category>
      <description>Google releases Gemini Omni 1.2 Flash with 15-second video, stronger consistency and a price of $0.07 per second for 720p video.</description>
    </item>
    <item>
      <title>New GPT-6-Cyber model in the API for organizations with Daybreak Red access</title>
      <link>https://zabio.ai/news/2026-11-05-openai-new-gpt-6-cyber-model-in-the-api-for-organizations-with/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-05-openai-new-gpt-6-cyber-model-in-the-api-for-organizations-with/</guid>
      <pubDate>Thu, 05 Nov 2026 17:10:24 GMT</pubDate>
      <category>openai</category>
      <description>OpenAI added the gpt-6-cyber model to the API. It needs Daybreak Red approval and costs $20.00 input and $100.00 output per 1M tokens.</description>
    </item>
    <item>
      <title>Anthropic adds an IP allowlist for the Claude Console, in beta</title>
      <link>https://zabio.ai/news/2026-11-05-anthropic-anthropic-adds-an-ip-allowlist-for-the-claude-console-in/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-05-anthropic-anthropic-adds-an-ip-allowlist-for-the-claude-console-in/</guid>
      <pubDate>Thu, 05 Nov 2026 16:45:24 GMT</pubDate>
      <category>anthropic</category>
      <description>Organizations can limit API requests to a list of IPv4 and IPv6 addresses or ranges, up to 1,000 per organization.</description>
    </item>
    <item>
      <title>OpenAI launches Images 2.6</title>
      <link>https://zabio.ai/news/2026-11-05-openai-openai-launches-images-2-6/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-05-openai-openai-launches-images-2-6/</guid>
      <pubDate>Thu, 05 Nov 2026 00:15:16 GMT</pubDate>
      <category>openai</category>
      <description>OpenAI released Images 2.6 with faster image generation, GIF creation, more precise editing, and new API controls.</description>
    </item>
    <item>
      <title>OpenAI releases GPT-Image-2.6 Sunburst and GPT-Image-2.6 Flare</title>
      <link>https://zabio.ai/news/2026-11-04-openai-openai-releases-gpt-image-2-6-sunburst-and-gpt-image-2-6-flare/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-04-openai-openai-releases-gpt-image-2-6-sunburst-and-gpt-image-2-6-flare/</guid>
      <pubDate>Wed, 04 Nov 2026 23:25:23 GMT</pubDate>
      <category>openai</category>
      <description>Two new image models that generate and edit images, with adjustable complexity, SVG output and animated GIF output.</description>
    </item>
    <item>
      <title>Together AI adds moonshotai/Kimi-K3.1-Pro to its serverless API</title>
      <link>https://zabio.ai/news/2026-11-04-together-together-ai-adds-moonshotai-kimi-k3-1-pro-to-its-serverless/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-04-together-together-ai-adds-moonshotai-kimi-k3-1-pro-to-its-serverless/</guid>
      <pubDate>Wed, 04 Nov 2026 17:05:21 GMT</pubDate>
      <category>together</category>
      <description>Together AI added the moonshotai/Kimi-K3.1-Pro model to its serverless API, priced at $2.00 input and $10.00 output per 1M tokens.</description>
    </item>
    <item>
      <title>xAI retires grok-imagine-image-quality on November 2, 2026</title>
      <link>https://zabio.ai/news/2026-11-04-xai-xai-retires-grok-imagine-image-quality-on-november-2-2026/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-04-xai-xai-retires-grok-imagine-image-quality-on-november-2-2026/</guid>
      <pubDate>Wed, 04 Nov 2026 07:30:29 GMT</pubDate>
      <category>xai</category>
      <description>Requests to grok-imagine-image-quality now go to grok-imagine-image-2.0 with quality set to low, at a lower price per image. No code change is required.</description>
    </item>
    <item>
      <title>Together AI adds Messages API and Responses API for GLM-5.2 and GLM-5.3</title>
      <link>https://zabio.ai/news/2026-11-03-together-together-ai-adds-messages-api-and-responses-api-for-glm-5-2/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-03-together-together-ai-adds-messages-api-and-responses-api-for-glm-5-2/</guid>
      <pubDate>Tue, 03 Nov 2026 21:30:21 GMT</pubDate>
      <category>together</category>
      <description>Together AI now supports the Anthropic Messages API and the OpenAI Responses API, and GLM-5.2 and GLM-5.3 support both.</description>
    </item>
    <item>
      <title>Gemini 4 Argon rolls out globally in the Gemini API at an introductory price</title>
      <link>https://zabio.ai/news/2026-11-02-google-gemini-4-argon-rolls-out-globally-in-the-gemini-api-at-an/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-11-02-google-gemini-4-argon-rolls-out-globally-in-the-gemini-api-at-an/</guid>
      <pubDate>Mon, 02 Nov 2026 17:55:34 GMT</pubDate>
      <category>google</category>
      <description>Google's Gemini 4 Argon, a frontier reasoning model, is now available in the Gemini API and other Google products at an introductory price.</description>
    </item>
    <item>
      <title>Anthropic deprecates fast mode for Claude Opus 4.8</title>
      <link>https://zabio.ai/news/2026-10-30-anthropic-anthropic-deprecates-fast-mode-for-claude-opus-4-8/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-30-anthropic-anthropic-deprecates-fast-mode-for-claude-opus-4-8/</guid>
      <pubDate>Fri, 30 Oct 2026 21:10:28 GMT</pubDate>
      <category>anthropic</category>
      <description>Fast mode for Claude Opus 4.8 is deprecated and will be removed on November 30, 2026.</description>
    </item>
    <item>
      <title>OpenAI releases GPT-6 Rosalind for approved life sciences research</title>
      <link>https://zabio.ai/news/2026-10-30-openai-openai-releases-gpt-6-rosalind-for-approved-life-sciences/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-30-openai-openai-releases-gpt-6-rosalind-for-approved-life-sciences/</guid>
      <pubDate>Fri, 30 Oct 2026 17:40:22 GMT</pubDate>
      <category>openai</category>
      <description>OpenAI released GPT-6 Rosalind, a biology model, for organizations approved through its trusted-access program.</description>
    </item>
    <item>
      <title>Moonshot Batch API now supports kimi-k3.1-pro and kimi-k3</title>
      <link>https://zabio.ai/news/2026-10-30-moonshot-moonshot-batch-api-now-supports-kimi-k3-1-pro-and-kimi-k3/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-30-moonshot-moonshot-batch-api-now-supports-kimi-k3-1-pro-and-kimi-k3/</guid>
      <pubDate>Fri, 30 Oct 2026 12:15:42 GMT</pubDate>
      <category>moonshot</category>
      <description>The Moonshot Batch API now accepts kimi-k3.1-pro and kimi-k3 on both the CN and Global platforms.</description>
    </item>
    <item>
      <title>Computer Use tool now generally available in the Gemini API</title>
      <link>https://zabio.ai/news/2026-10-30-google-computer-use-tool-now-generally-available-in-the-gemini-api/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-30-google-computer-use-tool-now-generally-available-in-the-gemini-api/</guid>
      <pubDate>Fri, 30 Oct 2026 12:15:23 GMT</pubDate>
      <category>google</category>
      <description>Google made the Computer Use tool generally available across the Interactions and GenerateContent APIs.</description>
    </item>
    <item>
      <title>Claude API raises request size limit to 128 MB for Messages and Token Counting</title>
      <link>https://zabio.ai/news/2026-10-29-anthropic-claude-api-raises-request-size-limit-to-128-mb-for/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-29-anthropic-claude-api-raises-request-size-limit-to-128-mb-for/</guid>
      <pubDate>Thu, 29 Oct 2026 17:30:19 GMT</pubDate>
      <category>anthropic</category>
      <description>Anthropic raised the maximum request size for the Messages API and the Token Counting API from 32 MB to 128 MB.</description>
    </item>
    <item>
      <title>The weights of Mistral Large 4 (Le Chonk) are now open</title>
      <link>https://zabio.ai/news/2026-10-29-mistral-the-weights-of-mistral-large-4-le-chonk-are-now-open/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-29-mistral-the-weights-of-mistral-large-4-le-chonk-are-now-open/</guid>
      <pubDate>Thu, 29 Oct 2026 14:05:35 GMT</pubDate>
      <category>mistral</category>
      <description>The weights of Mistral Large 4 (Le Chonk) are now open. The model is priced at $0.70 input and $2.10 output per million tokens.</description>
    </item>
    <item>
      <title>OpenAI releases GPT-6.1 Terra for visual reasoning and computer use</title>
      <link>https://zabio.ai/news/2026-10-26-openai-openai-releases-gpt-6-1-terra-for-visual-reasoning-and/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-26-openai-openai-releases-gpt-6-1-terra-for-visual-reasoning-and/</guid>
      <pubDate>Mon, 26 Oct 2026 17:55:24 GMT</pubDate>
      <category>openai</category>
      <description>GPT-6.1 Terra accepts text and image input, has a 1,050,000-token context window and costs $1.00 per 1M input tokens.</description>
    </item>
    <item>
      <title>Together adds MiniMaxAI/MiniMax-M3.1-Flash</title>
      <link>https://zabio.ai/news/2026-10-23-together-together-adds-minimaxai-minimax-m3-1-flash/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-23-together-together-adds-minimaxai-minimax-m3-1-flash/</guid>
      <pubDate>Fri, 23 Oct 2026 04:15:19 GMT</pubDate>
      <category>together</category>
      <description>Together lists a new serverless model with a 1,048,576 context length and prices for input, output, and cached input.</description>
    </item>
    <item>
      <title>Anthropic launches Claude API access through Google Cloud</title>
      <link>https://zabio.ai/news/2026-10-22-anthropic-anthropic-launches-claude-api-access-through-google-cloud/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-22-anthropic-anthropic-launches-claude-api-access-through-google-cloud/</guid>
      <pubDate>Thu, 22 Oct 2026 20:50:28 GMT</pubDate>
      <category>anthropic</category>
      <description>Developers can access the Claude API through a Google Cloud project, with Google Cloud IAM controlling access and Marketplace handling billing.</description>
    </item>
    <item>
      <title>OpenAI makes the Decisions API generally available</title>
      <link>https://zabio.ai/news/2026-10-22-openai-openai-makes-the-decisions-api-generally-available/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-22-openai-openai-makes-the-decisions-api-generally-available/</guid>
      <pubDate>Thu, 22 Oct 2026 18:45:27 GMT</pubDate>
      <category>openai</category>
      <description>The Decisions API is generally available with gpt-6-luna, with shared input caching and support for up to 400 questions per request.</description>
    </item>
    <item>
      <title>Z.ai GLM 5.3 Flash is now available</title>
      <link>https://zabio.ai/news/2026-10-22-mistral-z-ai-glm-5-3-flash-is-now-available/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-22-mistral-z-ai-glm-5-3-flash-is-now-available/</guid>
      <pubDate>Thu, 22 Oct 2026 18:15:31 GMT</pubDate>
      <category>mistral</category>
      <description>Mistral lists Z.ai GLM 5.3 Flash with a 1M-token context window and prices of $0.15/M input tokens and $0.50/M output tokens.</description>
    </item>
    <item>
      <title>OpenAI: choose which service tiers projects can use in Platform settings</title>
      <link>https://zabio.ai/news/2026-10-21-openai-openai-choose-which-service-tiers-projects-can-use-in/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-21-openai-openai-choose-which-service-tiers-projects-can-use-in/</guid>
      <pubDate>Wed, 21 Oct 2026 20:27:21 GMT</pubDate>
      <category>openai</category>
      <description>OpenAI now lets you restrict by project and organization which service tiers requests use; requests that resolve to a disallowed tier get a 400 error.</description>
    </item>
    <item>
      <title>Anthropic releases API version 2026-10-21 and updated SDKs</title>
      <link>https://zabio.ai/news/2026-10-21-anthropic-anthropic-releases-api-version-2026-10-21-and-updated-sdks/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-21-anthropic-anthropic-releases-api-version-2026-10-21-and-updated-sdks/</guid>
      <pubDate>Wed, 21 Oct 2026 17:51:57 GMT</pubDate>
      <category>anthropic</category>
      <description>Anthropic released API version 2026-10-21 with breaking changes to sampling parameters and message formats.</description>
    </item>
    <item>
      <title>OpenAI adds 12-hour prompt caching to the API</title>
      <link>https://zabio.ai/news/2026-10-20-openai-openai-adds-12-hour-prompt-caching-to-the-api/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-20-openai-openai-adds-12-hour-prompt-caching-to-the-api/</guid>
      <pubDate>Tue, 20 Oct 2026 17:55:25 GMT</pubDate>
      <category>openai</category>
      <description>OpenAI now lets developers choose a 12-hour cache lifetime in the Responses API for GPT-5.6 and later.</description>
    </item>
    <item>
      <title>DeepSeek V4.1 Flash-1017 now available as a serverless model on Together AI</title>
      <link>https://zabio.ai/news/2026-10-19-together-deepseek-v4-1-flash-1017-now-available-as-a-serverless/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-19-together-deepseek-v4-1-flash-1017-now-available-as-a-serverless/</guid>
      <pubDate>Mon, 19 Oct 2026 15:40:20 GMT</pubDate>
      <category>together</category>
      <description>Together AI added a serverless DeepSeek model with a 1,000,000 context length and priced input and output per 1M tokens.</description>
    </item>
    <item>
      <title>DeepSeek opens up its in-house search API</title>
      <link>https://zabio.ai/news/2026-10-17-deepseek-deepseek-opens-up-its-in-house-search-api/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-17-deepseek-deepseek-opens-up-its-in-house-search-api/</guid>
      <pubDate>Sat, 17 Oct 2026 08:10:28 GMT</pubDate>
      <category>deepseek</category>
      <description>DeepSeek opened up its in-house search API, powered by DeepSeek Search, which indexes hundreds of billions of web pages and supports multilingual search.</description>
    </item>
    <item>
      <title>OpenAI releases prompt-to-voice for all API customers</title>
      <link>https://zabio.ai/news/2026-10-16-openai-openai-releases-prompt-to-voice-for-all-api-customers/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-16-openai-openai-releases-prompt-to-voice-for-all-api-customers/</guid>
      <pubDate>Fri, 16 Oct 2026 16:50:26 GMT</pubDate>
      <category>openai</category>
      <description>Developers can create a custom voice from a text description through the API or voice Playground and reuse it in GPT-Live sessions; nine new voices were also added.</description>
    </item>
    <item>
      <title>Together adds serverless model reflection/Beam-1.0-501B-A23B</title>
      <link>https://zabio.ai/news/2026-10-16-together-together-adds-serverless-model-reflection-beam-1-0-501b-a23b/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-16-together-together-adds-serverless-model-reflection-beam-1-0-501b-a23b/</guid>
      <pubDate>Fri, 16 Oct 2026 14:45:34 GMT</pubDate>
      <category>together</category>
      <description>The new model has a context length of 1,048,576, NVFP4 quantization, function calling, structured outputs, and costs $0.40 per 1M input tokens and $1.20 per 1M output tokens.</description>
    </item>
    <item>
      <title>OpenAI adds native Decisions integration in GPT-Live</title>
      <link>https://zabio.ai/news/2026-10-16-openai-openai-adds-native-decisions-integration-in-gpt-live/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-16-openai-openai-adds-native-decisions-integration-in-gpt-live/</guid>
      <pubDate>Fri, 16 Oct 2026 00:35:41 GMT</pubDate>
      <category>openai</category>
      <description>GPT-Live now supports native Decisions integration, letting apps act while the user is speaking and react to what Live says.</description>
    </item>
    <item>
      <title>Fast mode for Claude Opus 5.5 leaves research preview</title>
      <link>https://zabio.ai/news/2026-10-15-anthropic-fast-mode-for-claude-opus-5-5-leaves-research-preview/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-15-anthropic-fast-mode-for-claude-opus-5-5-leaves-research-preview/</guid>
      <pubDate>Thu, 15 Oct 2026 19:55:56 GMT</pubDate>
      <category>anthropic</category>
      <description>Claude Opus 5.5 fast mode no longer needs an access request or the fast-mode-2026-02-01 beta header, and now has a default rate limit in every usage tier.</description>
    </item>
    <item>
      <title>Meta Muse Spark Open is live on Together AI</title>
      <link>https://zabio.ai/news/2026-10-15-together-meta-muse-spark-open-is-live-on-together-ai/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-15-together-meta-muse-spark-open-is-live-on-together-ai/</guid>
      <pubDate>Thu, 15 Oct 2026 18:31:41 GMT</pubDate>
      <category>together</category>
      <description>Together AI now serves Muse Spark Open, Meta's open-weight model for agentic coding, with a 1M-token context window on serverless and dedicated infrastructure.</description>
    </item>
    <item>
      <title>New serverless model nvidia/NVIDIA-Nemotron-3.5-Super-VL with 262,144 context</title>
      <link>https://zabio.ai/news/2026-10-15-together-new-serverless-model-nvidia-nvidia-nemotron-3-5-super-vl/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-15-together-new-serverless-model-nvidia-nvidia-nemotron-3-5-super-vl/</guid>
      <pubDate>Thu, 15 Oct 2026 15:42:20 GMT</pubDate>
      <category>together</category>
      <description>Together now serves nvidia/NVIDIA-Nemotron-3.5-Super-VL serverless with 262,144 context, image input, function calling and structured outputs.</description>
    </item>
    <item>
      <title>Mistral introduces Timestral time-series foundation models</title>
      <link>https://zabio.ai/news/2026-10-15-mistral-mistral-introduces-timestral-time-series-foundation-models/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-15-mistral-mistral-introduces-timestral-time-series-foundation-models/</guid>
      <pubDate>Thu, 15 Oct 2026 15:07:00 GMT</pubDate>
      <category>mistral</category>
      <description>Mistral released Timestral, a family of four time-series foundation models with up to 209M parameters.</description>
    </item>
    <item>
      <title>xAI deprecates grok-voice-2.0, upgrade to grok-voice-3.0 at the same price</title>
      <link>https://zabio.ai/news/2026-10-14-xai-xai-deprecates-grok-voice-2-0-upgrade-to-grok-voice-3-0-at-the/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-14-xai-xai-deprecates-grok-voice-2-0-upgrade-to-grok-voice-3-0-at-the/</guid>
      <pubDate>Wed, 14 Oct 2026 16:33:17 GMT</pubDate>
      <category>xai</category>
      <description>xAI deprecated grok-voice-2.0 and directs developers to grok-voice-3.0 or grok-voice-latest at the same price.</description>
    </item>
    <item>
      <title>xAI: Grok Voice 3.0 is available with Speech to Speech</title>
      <link>https://zabio.ai/news/2026-10-14-xai-xai-grok-voice-3-0-is-available-with-speech-to-speech/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-14-xai-xai-grok-voice-3-0-is-available-with-speech-to-speech/</guid>
      <pubDate>Wed, 14 Oct 2026 16:32:13 GMT</pubDate>
      <category>xai</category>
      <description>grok-voice-3.0 is now available with Speech to Speech at the same price as Grok Voice 2.0, and the alias grok-voice-latest points to it.</description>
    </item>
    <item>
      <title>Grok Voice 1.0 is available with the Speech to Speech API</title>
      <link>https://zabio.ai/news/2026-10-14-xai-grok-voice-1-0-is-available-with-the-speech-to-speech-api/</link>
      <guid isPermaLink="true">https://zabio.ai/news/2026-10-14-xai-grok-voice-1-0-is-available-with-the-speech-to-speech-api/</guid>
      <pubDate>Wed, 14 Oct 2026 16:30:35 GMT</pubDate>
      <category>xai</category>
      <description>xAI's Grok Voice 1.0 is now available with the Speech to Speech API.</description>
    </item>
  </channel>
</rss>
