# Ollama changelog > Get up and running with large language models locally. - Vendor: Ollama - Category: AI - Official site: https://ollama.com - Tracked by: What's New (https://whatsnew.fyi/product/ollama) - Harvested from: GitHub (ollama/ollama) - Entries below: 12 (newest first) What's New is an index, not a publisher: every entry below links to the vendor's own release notes, which are the authoritative source. Entries are labelled where they are hand-curated sample data, pre-releases, or drawn from a secondary source such as a developer blog. ## Releases ### v0.32.5 - Date: 2026-07-27 - Version: v0.32.5 - Original notes: https://github.com/ollama/ollama/releases/tag/v0.32.5 - Permalink: https://whatsnew.fyi/product/ollama/releases/v0.32.5-rc0 - **fixed** — Fixed an MLX Metal bug that could reduce output quality for NVFP4 models, particularly Laguna ##### What's Changed * Fixed an MLX Metal bug that could reduce output quality for NVFP4 models, particularly Laguna. **Full Changelog**: https://github.com/ollama/ollama/compare/v0.32.4...v0.32.5 ### v0.32.4 - Date: 2026-07-25 - Version: v0.32.4 - Original notes: https://github.com/ollama/ollama/releases/tag/v0.32.4 - Permalink: https://whatsnew.fyi/product/ollama/releases/v0.32.4 - **added** — Support Laguna on Apple GPUs via the MLX engine - **changed** — Quantize draft-model output heads at the requested type when creating speculative-decoding drafts - **fixed** — Fixed Qwen3 MoE decoding for differently-quantized experts and improved packed gate/up projection performance - **changed** — Faster packed gate/up projection for Qwen3 MoE decoding on M5 Max ##### What's Changed - Support Laguna on Apple GPUs via the MLX engine - Quantize draft-model output heads at the requested type when creating speculative-decoding drafts. - Fixed Qwen3 MoE decoding for differently-quantized experts, plus faster packed gate/up projection (~4–9% on M5 Max). **Full Changelog**: https://github.com/ollama/ollama/compare/v0.32.3...v0.32.4 ### v0.32.3 - Date: 2026-07-23 - Version: v0.32.3 - Original notes: https://github.com/ollama/ollama/releases/tag/v0.32.3 - Permalink: https://whatsnew.fyi/product/ollama/releases/v0.32.3 - **fixed** — Fixed model downloads that stall before sending data - **fixed** — Restored Claude Code Channels in integrations - **fixed** — Fixed Anthropic thinking streams in integrations - **changed** — Made Hermes Desktop respect --force-build flag - **added** — Added CUDA support on Windows ARM64 - **added** — Added B200 GPU support through CUDA 12 - **changed** — Reduced memory usage on Linux CUDA/ROCm iGPUs - **added** — Added chat, thinking, and tool calling support for Laguna 2.1 models - **fixed** — Fixed Metal inference for Laguna 2.1 models - **fixed** — Fixed GLM tool calls being silently dropped at the end of generation - **changed** — Updated the MLX engine - **changed** — Updated the llama.cpp engine ##### What's Changed - Fixed model downloads that stall before sending data. - Improved integrations: restored Claude Code Channels, fixed Anthropic thinking streams, and made Hermes Desktop respect `--force-build`. - Expanded GPU support with CUDA on Windows ARM64, B200 support through CUDA 12, and lower memory use on Linux CUDA/ROCm iGPUs. - Added chat, thinking, and tool calling support for Laguna 2.1 models, including a Metal inference fix. - Fixed GLM tool calls being silently dropped at the end of generation. - Updated the MLX and llama.cpp engines. **Full Changelog**: https://github.com/ollama/ollama/compare/v0.32.1...v0.32.3 ### v0.32.2 - Date: 2026-07-20 - Version: v0.32.2 - Original notes: https://github.com/ollama/ollama/releases/tag/v0.32.2 - Permalink: https://whatsnew.fyi/product/ollama/releases/v0.32.2 - Labels: Pre-release Withdrawn, please use 0.32.3 or newer ### v0.32.1 - Date: 2026-07-16 - Version: v0.32.1 - Original notes: https://github.com/ollama/ollama/releases/tag/v0.32.1 - Permalink: https://whatsnew.fyi/product/ollama/releases/v0.32.1 - **changed** — Improved Gemma 4 tool calling and multi-turn reasoning, including more reliable tool-response continuations - **fixed** — Fixed a recurrent MLX model cache leak that could increase memory use across requests - **changed** — Improved MLX cache snapshot performance - **changed** — MLX text model loading now respects OLLAMA_LOAD_TIMEOUT - **changed** — Agent web search and fetch now tell users to run ollama signin when authentication is required - **changed** — The interactive agent now receives the current working directory for better project context - **fixed** — Fixed ollama launch so choosing Pick another model for a deprecated model passed with --model opens the model picker ##### What's Changed - Improved Gemma 4 tool calling and multi-turn reasoning, including more reliable tool-response continuations - Fixed a recurrent MLX model cache leak that could increase memory use across requests, and improved cache snapshot performance - MLX text model loading now respects `OLLAMA_LOAD_TIMEOUT` - Agent web search and fetch now tell users to run `ollama signin` when authentication is required - The interactive agent now receives the current working directory for better project context - Fixed `ollama launch` so choosing **Pick another model** for a deprecated model passed with `--model` opens the model picker - Updated VS Code setup documentation for the official Ollama extension **Full Changelog**: https://github.com/ollama/ollama/compare/v0.32.0...v0.32.1-rc0 ### v0.32.0 - Date: 2026-07-11 - Version: v0.32.0 - Original notes: https://github.com/ollama/ollama/releases/tag/v0.32.0 - Permalink: https://whatsnew.fyi/product/ollama/releases/v0.32.0 - **added** — New interactive agent experience when running ollama command to help with coding and delegating work - **changed** — Renamed Codex App integration to ChatGPT with ollama launch chatgpt command - **changed** — Simplified integration selection in ollama launch menu to show only the most popular integrations - **added** — Deprecation warnings before launching older agent models including CodeLlama, Qwen2.5(-coder), Llama 3.x, Mistral, StarCoder, and base DeepSeek-R1 tags ##### What's Changed - New interactive agent experience: running `ollama` now launches an agent to help you code and delegate work ``` ❯ ollama Ollama 0.32.0 ▸ Chat, Code, & Work (glm-5.2:cloud) Chat with models, code, search the web, and delegate real work ``` - Renamed the Codex App integration to ChatGPT: use ollama launch chatgpt (and --restore to return to your usual ChatGPT profile) - Simplified integration selection: the ollama launch menu now only offers the most popular integrations (other integrations can be accessed through `ollama launch` - Warns before launching older agent models: CodeLlama, Qwen2.5(-coder), Llama 3.x, Mistral, StarCoder, and the base DeepSeek-R1 tags now prompt a deprecation warning before ollama launch continues **Full Changelog**: https://github.com/ollama/ollama/compare/v0.31.2...v0.32.0 ### v0.31.2 - Date: 2026-07-06 - Version: v0.31.2 - Original notes: https://github.com/ollama/ollama/releases/tag/v0.31.2 - Permalink: https://whatsnew.fyi/product/ollama/releases/v0.31.2 - **changed** — Enabled flash attention on older NVIDIA GPUs (compute capability 6.x) - **changed** — iGPU can now offload vision models with padding to fit available memory - **fixed** — Fixed structured output for thinking models when thinking is disabled - **changed** — Hardened GGUF model creation - **changed** — ollama launch for Claude Code now disables telemetry by default - **fixed** — Fixed loading models on paths with non-UTF-8 characters - **changed** — Updated the MLX and llama.cpp engines ##### What's Changed * Enabled flash attention on older NVIDIA GPUs (compute capability 6.x) * iGPU can now offload vision models with padding to fit available memory * Fixed structured output for thinking models when thinking is disabled * Hardened GGUF model creation * `ollama launch` for Claude Code now disables telemetry by default * Fixed loading models on paths with non-UTF-8 characters * Updated the MLX and llama.cpp engines ##### New Contributors * @kevinpark1217 made their first contribution in https://github.com/ollama/ollama/pull/16949 **Full Changelog**: https://github.com/ollama/ollama/compare/v0.31.1...v0.31.2 ### v0.31.1 - Date: 2026-06-30 - Version: v0.31.1 - Original notes: https://github.com/ollama/ollama/releases/tag/v0.31.1 - Permalink: https://whatsnew.fyi/product/ollama/releases/v0.31.1 - **changed** — Improved Gemma 4 multi-token prediction (MTP) performance on Apple Silicon - **changed** — Tightened Gemma 4 MoE model loading in the MLX engine - **changed** — Updated the MLX engine to the latest version, including a new small-batch matmul kernel - **changed** — Updated the underlying llama.cpp engine to build 9840 ##### Faster Gemma 4 on Apple Silicon Screenshot 2026-06-30 at 5 25 29 PM Gemma 4 is now significantly faster in Ollama on Apple Silicon, generating tokens nearly 90% faster on average across a coding-agent benchmark by leveraging multi-token prediction (MTP). Ollama auto-tunes how many tokens to draft as it runs, so the speedup is on by default, requires no configuration, and does not change the model's output. ##### What's Changed - Tightened Gemma 4 MoE model loading in the MLX engine - Updated the MLX engine to the latest version, including a new small-batch matmul kernel - Updated the underlying llama.cpp engine to build 9840 - Improved Gemma 4 multi-token prediction (MTP) performance **Full Changelog**: https://github.com/ollama/ollama/compare/v0.30.12...v0.31.1 ### v0.30.12-rc0 — v0.30.12 - Date: 2026-06-29 - Version: v0.30.12-rc0 - Original notes: https://github.com/ollama/ollama/releases/tag/v0.30.12-rc0 - Permalink: https://whatsnew.fyi/product/ollama/releases/v0.30.12-rc0 - Labels: Pre-release - **fixed** — Ignore braces inside JSON strings when detecting tool call end - **changed** — Bump mlx dependency - **changed** — Update llama.cpp ##### What's Changed * tools: ignore braces inside JSON strings when detecting tool call end by @aditya-786 in https://github.com/ollama/ollama/pull/16937 * mlx: bump dependency by @dhiltgen in https://github.com/ollama/ollama/pull/16935 * llama.cpp update by @dhiltgen in https://github.com/ollama/ollama/pull/16960 ##### New Contributors * @aditya-786 made their first contribution in https://github.com/ollama/ollama/pull/16937 **Full Changelog**: https://github.com/ollama/ollama/compare/v0.30.11...v0.30.12-rc0 ### v0.30.11 - Date: 2026-06-25 - Version: v0.30.11 - Original notes: https://github.com/ollama/ollama/releases/tag/v0.30.11 - Permalink: https://whatsnew.fyi/product/ollama/releases/v0.30.11 - **added** — Add thinking capability detection to opencode - **added** — Auto-install Claude Code - **added** — Auto-install opencode when missing - **fixed** — Fix inverted iGPU/dGPU Vulkan classification on Windows hybrid graphics - **changed** — Unify and tune speculative decoding - **changed** — Detect model drift when Codex App UI switches - **added** — Add sm_86 architecture to cuda_v13_windows preset - **changed** — Size mmproj offload by projector memory - **changed** — Preserve generation headroom for shifted prompts - **changed** — Default qwen2.5vl window attention metadata - **changed** — Use host Vulkan loader on Windows - **changed** — Update and fix CUDA JIT packaging - **fixed** — Fix ollama ps double-counting mmap'd weights on partial offload - **changed** — Align generate with native chat templates - **added** — Add CC 87 for CUDA v13 on Jetson - **changed** — Update llama.cpp version ##### What's Changed * launch: add thinking capability detection to opencode by @hoyyeva in https://github.com/ollama/ollama/pull/15434 * launch: auto-install Claude Code by @hoyyeva in https://github.com/ollama/ollama/pull/16802 * launch: auto-install opencode when missing by @hoyyeva in https://github.com/ollama/ollama/pull/16806 * discover: fix inverted iGPU/dGPU Vulkan classification on Windows hybrid graphics by @Sahil170595 in https://github.com/ollama/ollama/pull/16669 * mlxrunner: unify and tune speculative decoding by @jessegross in https://github.com/ollama/ollama/pull/16791 * launch/codex: detect model drift when Codex App UI switches by @BruceMacD in https://github.com/ollama/ollama/pull/16864 * llama: add sm_86 architecture to cuda_v13_windows preset by @anishesg in https://github.com/ollama/ollama/pull/16834 * llm: size mmproj offload by projector memory by @dhiltgen in https://github.com/ollama/ollama/pull/16866 * docs: document max think level by @ParthSareen in https://github.com/ollama/ollama/pull/16877 * llm: preserve generation headroom for shifted prompts by @ParthSareen in https://github.com/ollama/ollama/pull/16856 * llama: default qwen2.5vl window attention metadata by @dhiltgen in https://github.com/ollama/ollama/pull/16868 * llm: use host Vulkan loader on Windows by @dhiltgen in https://github.com/ollama/ollama/pull/16869 * mlx: update and fix CUDA JIT packaging by @dhiltgen in https://github.com/ollama/ollama/pull/16871 * llm: fix ollama ps double-counting mmap'd weights on partial offload by @discobot in https://github.com/ollama/ollama/pull/16709 * docs: redesign docs landing and integrations overview by @hoyyeva in https://github.com/ollama/ollama/pull/16807 * server: align generate with native chat templates by @dhiltgen in https://github.com/ollama/ollama/pull/16878 * jetson: add CC 87 for CUDA v13 by @dhiltgen in https://github.com/ollama/ollama/pull/16628 * llama.cpp version update by @dhiltgen in https://github.com/ollama/ollama/pull/16548 ##### New Contributors * @Sahil170595 made their first contribution in https://github.com/ollama/ollama/pull/16669 * @anishesg made their first contribution in https://github.com/ollama/ollama/pull/16834 * @discobot made their first contribution in https://github.com/ollama/ollama/pull/16709 **Full Changelog**: https://github.com/ollama/ollama/compare/v0.30.10...v0.30.11-rc0 ### v0.30.10 - Date: 2026-06-17 - Version: v0.30.10 - Original notes: https://github.com/ollama/ollama/releases/tag/v0.30.10 - Permalink: https://whatsnew.fyi/product/ollama/releases/v0.30.10 - **added** — Command A and North family models now run on Apple Silicon with the MLX engine - **changed** — Update the underlying llama.cpp engine to build 9672 - **fixed** — Fix build artifacts for MLX ##### What's Changed * Command A and North family models now run on Apple Silicon with the MLX engine * Updated the underlying llama.cpp engine to build 9672 * Fixed build artifacts for MLX **Full Changelog**: https://github.com/ollama/ollama/compare/v0.30.9...v0.30.10 ### v0.30.9 - Date: 2026-06-15 - Version: v0.30.9 - Original notes: https://github.com/ollama/ollama/releases/tag/v0.30.9 - Permalink: https://whatsnew.fyi/product/ollama/releases/v0.30.9 - **added** — Support for Cohere2Moe architecture - **fixed** — Fixed LFM2 parser/render for cases where thinking was not emitted - **fixed** — Fixed issue where ollama launch claude and other coding agent or assistant use cases would only output one token - **changed** — Ollama will now return an error if a single message is larger than the current context window ##### What's Changed * Support for Cohere2Moe architecture * Fixed LFM2 parser/render for cases where thinking was not emitted * Fixed issue where `ollama launch claude` and other coding agent or assistant use cases would only output one token * Ollama will now return an error if a single message is larger than the current context window **Full Changelog**: https://github.com/ollama/ollama/compare/v0.30.8...v0.30.9-rc1