Private · Palo Alto, US · 22 releases · 2 this month
MiMo V2.6 Xiaomi MiMo's mimo-v2.6, a 9B model built for code and agent workflows, is now supported in Ollama. sh ollama run mimo-v2.6 AutoTrust decision models AutoTrust decision models read text and return choices, yes or no probabilities, and ratings for the questions you supply.
Ollama 0.40.13 now supports Liquid AI's LFM3 models, including text-to-speech capabilities, on Linux and Windows in addition to Apple silicon.
Ollama 0.40.12 adds support for the Liquid AI LFM3 model family, including text, vision, audio, and omni variants. The release also includes performance improvements for Apple silicon and updated MLX and llama.cpp dependencies.
Ollama now supports speech generation on macOS and introduces d3 decision models for classification and scoring tasks. The d3 family includes models ranging from 2B to 27B parameters.
Ollama now supports the NVIDIA Nemotron 3.5 Bolt E4B vision-language model, which features a 128K-token context window.
Ollama 0.40.9 adds speech generation on Apple Silicon, introduces the d3 decision model family, and expands Flint support to Linux and Windows. It also includes improvements to streaming responses and context length handling.
Ollama release 0.40.8 adds support for Recursive's Flint coding model, an Apache 2.0-licensed 40B-parameter MoE model with a 262K-token context window optimized for repository-level coding tasks.
- pplx-decider is a 27B decision model from Perplexity, fine-tuned from Qwen3.8-27B - Continuing an OpenAI-compatible Responses conversation no longer fails when switching between models with and without thinking, or when the history contains another provider's compaction items - Fixed streaming Responses requests where client tool calls (e.g.
- win: prevent orphaned runners - gguf: add the q2_0 tensor and file types - Launch clef in Ollama's cloud - use identifiers for --ask mode question keys
NVIDIA Nemotron 3.5 Super NVIDIA Nemotron 3.5 Super is a vision-language model for coding, reasoning, tool use, and document understanding. It pairs a 120B-parameter hybrid Mamba-2/Transformer mixture-of-experts backbone (12B active per token) with a C-RADIO vision encoder, and supports tool calling, an on/off thinking mode, and context up to 1M tokens.
Decision models in the CLI Decision models can now be used directly from the terminal with ollama run , ask yes/no questions, classify text into labels, or score against a rubric: ollama run nimble --ask "Is a refund being requested?" < ticket.txt ## - Kev decision models are now supported by the MLX runner on Apple silicon - ollama launch can now be used...
- ollama launch claude no longer shows a warning about Claude connectors - ollama launch codex no longer warns about service tiers - Paused background model upgrade for embedding and other request performance,
Model upgrades Models downloaded with earlier versions of Ollama are upgraded in the background the first time you run them, for better performance and compatibility when running on llama.cpp. To make downgrading safe, Ollama keeps the original copy as a backup, so upgraded models are temporarily kept on disk.
- server: proxy cloud usage and balance APIs - llama: fix clef head reads past 2GiB on windows - cmd: remove account step from CLI onboarding - manifest: avoid symlinks on Windows - docs: fix 6 dead links in README community integrations list - docs: fix broken download links in app README - mlx: drop carried metal residency patch now that it is upstream
Clef decision models Ollama now supports Clef and Clef Flash, Cloudflare's new open-source decision models, through /v1/systemone. Clef (27B) and Clef Flash (9B) are multimodal: requests can now include images alongside the text state, shared by all questions and scored jointly with it.
Decision models Ollama now supports decision models through /v1/systemone, based on TypeSafe’s Jev API. Decision models return choices, probabilities, and scores instead of text. Use them for tasks such as ticket triage, model routing, and content classification.
Models run on MLX on Apple Silicon by default In this release, on Apple Silicon devices, model architectures supported by the MLX runtime will automatically run on MLX.
- Structured outputs on thinking models now apply in a single pass, making them faster and more reliable. Fixed intermittent "model not found" errors with a large local library Fixed the macOS app becoming unresponsive when checking if ChatGPT or Codex is running. Qwen 3.8 prompt processing is faster on Apple Silicon.
GET /api/show now advertises each model's thinking controls and default: Available in the CLI with: ollama show gemma4 thinking levels false, true default true Available in the API with: curl http://localhost:11434/api/show -d ' {"model": "glm-5.3-flash:cloud"} ' { "thinking" : { "values" : [ " low " , " high " , " max " ], "default" : " max " } } Also...
- Added first-run setup when running ollama , with options to sign in or continue locally. Setup completion is shared with the desktop app on macOS and Windows. Added ollama://apps to open the desktop app’s Apps page directly on macOS and Windows. Fixed excessive memory growth during long generations with MLX speculative decoding. Updated llama.cpp.
- MLX safetensors ollama create no longer experimental. GGUF model creation now requires using llama.cpp tooling for safetensor conversion and quantization. - Improved MLX memory handling on Apple Silicon - Runaway repeat token detection now requires 100 repeat tokens for reduced false positives (e.g.
Open-source tool for running open-weight AI models locally, plus a paid cloud service.
Releasesby Glyph