Local LLMUpdates
Search

Release tracker

Verified local LLM release changes

Scan exact versions, dates, change types, verification state, and primary evidence in one place.

7 tracker records

Release date · evidence state · direct sources

Verified means checked against the cited source. Vendor performance claims are attributed, not independently reproduced.

  1. Date
    #
    Released
    Releasev0.33.2macOS app and Claude Desktop proxy fixesRelease tracker
    Entities
    • Ollama
    Changes
    • Release
    VerificationVerifiedChecked
    BreakingUnknown
    Summary

    Ollama restores following the system appearance, hands the macOS app off to an already-running instance, and prevents model-catalog updates from interrupting in-flight Claude Desktop proxy requests.

    Evidence
    1. Ollama v0.33.2 release notesOllama
  2. Date
    #
    Released
    Releasev0.33.0Claude Desktop gateway, caching, and packaging fixesRelease tracker
    Entities
    • Ollama
    Changes
    • Release
    • Performance
    • Compatibility
    VerificationVerifiedChecked
    BreakingUnknown
    Summary

    Ollama's v0.33.0 release lets Claude Desktop use Ollama as a third-party gateway provider, fixes prompt caching for recurrent-layer models, and fixes Linux and Windows packaging.

    Evidence
    1. v0.33.0Ollama
  3. Date
    #
    Released
    Releasev0.32.15Ollama reports v0.32.15 caching model metadata and normalizing Qwen 3.8 system messagesRelease tracker
    Entities
    • Ollama
    • Qwen 3.8
    Changes
    • Release
    • Performance
    • Compatibility
    VerificationVerifiedChecked
    BreakingUnknown
    Summary

    Ollama reports that v0.32.15 caches resolved model metadata, fixes chat and generate wedging after a parser error, and normalizes Qwen 3.8 system messages.

    Evidence
    1. v0.32.15Ollama
  4. Date
    #
    Released
    Releasev0.32.14WebP transcoding and Qwen system-message handlingRelease tracker
    Entities
    • Ollama
    Changes
    • Release
    • Compatibility
    VerificationVerifiedChecked
    BreakingUnknown
    Summary

    The official notes list WebP transcoding for llama-server and Qwen renderer tolerance for non-leading system messages.

    Evidence
    1. v0.32.14Ollama
  5. Date
    #
    Released
    Releasev0.32.12Qwen3.8-27B supportRelease tracker
    Entities
    • Qwen 3.8 27B
    • Ollama
    Changes
    • Release
    • Performance
    VerificationVerifiedChecked
    BreakingUnknown
    Summary

    The release adds Qwen3.8-27B support, with Ollama stating particular optimizations for Apple Silicon.

    Evidence
    1. v0.32.12Ollama
  6. Date
    #
    Released
    Releasev0.32.10NVFP4 prefill and repeat-penalty defaultsRelease tracker
    Entities
    • Ollama
    Changes
    • Release
    • Performance
    VerificationVerifiedChecked
    BreakingUnknown
    Summary

    Ollama reports about 7–8% faster prefill on the Qwen3.6 and Muse Glimmer NVFP4 MLX models with a global scale, changes the default to 1.0 instead of 1.1 for models without `repeat_penalty`, and fixes a shared-digest blob-verification case.

    Evidence
    1. v0.32.10Ollama
  7. Date
    #
    Released
    Releasev0.32.9Nemotron 3 architecture supportRelease tracker
    Entities
    • NVIDIA Nemotron 3.5 Lightning
    • Ollama
    Changes
    • Release
    VerificationVerifiedChecked
    BreakingUnknown
    Summary

    Ollama's release notes add the Nemotron 3 architecture and describe NVIDIA Nemotron 3.5 Lightning as an open 30B MoE model with 3B active parameters.

    Evidence
    1. v0.32.9Ollama