What’s new in open models

🔓 Chat over plain HTTP. Opening Open WebUI over plain HTTP at a LAN or Docker address now loads new and existing chats instead of staying on a spinner with "crypto.randomUUID is not a function". #33919

🦾 Coding agents as models. On an Open Terminal that offers agents, coding agents such as Claude Code, Codex and Pi now appear in the model picker. You can chat with them or call them by voice, choose their reasoning effort and service tier, and approve their tool requests in the

> [!TIP]

> [!TIP]

This release brings large performance fixes across Unsloth: faster typing and search in long chats, faster image/video loads, two-GPU image/video gen. It also adds separate browsers for each chat, floating browser window.

> [!TIP]

Build and deploy enterprise agents, workflows and apps faster with Mistral Studio and Le Chonk full control over your models, data and infrastructure.

This release introduces Qwen3.8-Flash-Next Dynamic V3.1, polishes the Glass themes (now in Beta), adds a Composer position slider and lets models interact with files in the built-in browser. It also brings lower idle memory usage, preserved chat context when switching GGUF models

This release adds OCP FP8 E4M3 quantization support, a new flash_attn_ext_rows op that makes the unified KV cache efficient for multi-sequence inference, and a generic batched tensor copy API. Four new architectures are supported (Clover, Limite, Metis-1.6 and Mistral Large 4), a

cosign verify \

> [!TIP]

Xiaomi MiMo's mimo-v2.6, a 9B model built for code and agent workflows, is now supported in Ollama.

Unsloth now has Glass themes and more customization options for backgrounds, greetings and chats. This release also brings faster generation, more multi-GPU training options and CPU training in Unsloth.

📑 Tabbed workspace. The Tabs button at the top of a chat now opens a workspace for files, multiple terminal sessions, artifacts, previews, voice calls and comments. Add tabs from the + menu, drag to reorder them, and create and save new files directly into the selected terminal.

What's Changed ollama launch claude now applies the model's output token limit for Claude Code, so cloud models no longer run into the 128k default cap Cloud web search responses on /api/chat are now decoded correctly, including compressed responses OpenAI-compatible endpoints…

1007 PRs from 257 contributors.

This release features 927 commits from 389 contributors (122 new)!

Make Unsloth your own with custom backgrounds and a improved UI/UX. This release also cuts app update times and speeds up Linux app and Qwen-Image gen when VRAM is limited. Also new: DeepSeek-V4.1-Flash-1017 in Unsloth Dynamic v3.1 GGUFs (preview), FLUX.2-klein-9B on low-RAM mach

The weights of Mistral Large 4 (Le Chonk) are now open. A 1T parameter multimodal model, and best open-weight model for coding, trained in Europe.

🧭 Analytics session explorer. Administrators can now open a Sessions tab in Admin Analytics that lists each chat with its tokens, responses, errors and time span, filters them by search, model, token range or sub-agents, and shows a chat's saved replies and tool calls with their

All LiteLLM Docker images are signed with cosign. This release is signed with the key introduced in commit e82a3c6.

Liquid AI's LFM3 models, including text-to-speech in lfm3-audio and lfm3-omni, now run on Linux and Windows as well as Apple silicon.

Halloween 👻 is here. Through October, a Trick or treat row in the sidebar opens eleven Halloween themes, spooky greetings and a few spiders and bats.

This release adds the LiquidAI LFM3 family (text, vision, audio and omni) together with a new experimental streaming-audio API in mtmd, and wires realtime audio in and out of the server through the new /v1/realtime and /tts endpoints. A batch of new architectures landed: DeepSeek

cosign verify \

cosign verify \

cosign verify \

cosign verify \

45 commits since v0.21.0 (October 16, 2026).

LFM3 is Liquid AI's new family of small models built for local agents, tool calling, and structured output, on Liquid's new tri-hybrid architecture. The family includes:

This release adds scheduled tasks, FLUX-3 edit support, perf improvements and more. Also new: MiniMax-M3.1-Flash, face touch-ups for images and faster performance + chats on Mac.

cosign verify \

cosign verify \

cosign verify \

cosign verify \

🎯 Load selected skills on demand. Administrators can turn on "Load selected skills on demand" under Admin Settings > General or set "ENABLE_SKILL_LAZY_LOADING", so skills selected in the chat or attached to a model are only listed for the model to load when it needs them instead

> [!IMPORTANT]

> [!IMPORTANT]

> [!IMPORTANT]

> [!IMPORTANT]

Ollama can now generate speech on macOS. ollama run plays text or a chat model's reply aloud as the audio is generated, and apps can stream audio through the OpenAI-compatible /v1/audio/speech endpoint.

Recursive's Flint, MOSS-Image and NVIDIA's Nemotron Bolt 4B are here! Run them locally in Unsloth. Also new: a chat-only Unsloth Desktop install, MLX MoE models bigger than your Mac's memory and an OpenAI-compatible fine-tuning API.

Nemotron 3.5 Bolt E4B is NVIDIA’s smaller, Gemma 4-based vision-language model for fast local conversations, image understanding, and tool use. It supports a 128K-token context window.

Flint on Linux and Windows Flint now runs on Linux and Windows as well as macOS. Other improvements Streaming OpenAI Responses and Anthropic Messages requests now report generation failures instead of ending silently after partial output. ollama launch now passes cloud models'…

Flint brings Recursive's coding model to Ollama. It is designed for the work coding agents do across a repository: understanding unfamiliar code, tracing bugs, implementing features, and adding tests. Reasoning, multi-turn tool use, and a 262K-token context window support longer

Mistral is collaborating with NVIDIA to develop a new generation of AI-native engineering workflows powered by NVIDIA accelerated computing.

All LiteLLM Docker images are signed with cosign. Every release is signed with the same key introduced in commit 0112e53.

cosign verify \

Voxtral 3: Mistral's voice agent SDK for real-time speech in 50+ languages — accurate transcription, emotional TTS, deep tool use. EU-hosted or self-deployed.

DeepSeek-V4.1-Flash-1017 and FLUX.3 Image Klein are here! Run both locally in Unsloth. Also new: images and videos in chat, multi-GPU training with FSDP2 (alpha) and one-command MCP setup for coding agents.

950 PRs from 245 contributors.

All LiteLLM Docker images are signed with cosign. Every release is signed with the same key introduced in commit 0112e53.

pplx-decider is a 27B decision model from Perplexity, fine-tuned from

v0.32.0 Highlights This release features 814 commits from 363 contributors (121 new)! Structured decisions and generic parsers : the new POST /v1/systemone endpoint answers choice , noul and score questions over a structured state ( #59299 ), is served by default ( #60677 ,…

What's Changed win: prevent orphaned runners by @dhiltgen in #18500 gguf: add the q2_0 tensor and file types by @jmorganca in #18967 Launch clef in Ollama's cloud by @jberg5 in #19036 use identifiers for --ask mode question keys by @jberg5 in #19041 Full Changelog :…

59 commits since v0.20.0 (September 30, 2026).

🔐 End-to-end encrypted chats. Direct messages and private group channels can now hold a temporary end-to-end encrypted chat: a member picks Encrypted Chat in the channel menu, the other members see "Encrypted chat in progress" with a Join link, messages of up to 16 KiB are encry

Fix attention value scaling in MiMo-V2-Flash by @ricky-chaoju in https://github.com/ml-explore/mlx-lm/pull/1971

fix muse template by @nastya236 in https://github.com/ml-explore/mlx-lm/pull/1977

Run NVIDIA's new Nemotron 3.5 models and Meta's new Muse Spark Open locally in Unsloth. New: text-diffusion LLMs, DeepSpeed training and better prompt caching for coding agents.

Honour the left-padding mask in MiniCPM3 and Phixtral batches by @bhaskargurram-ai in https://github.com/ml-explore/mlx-lm/pull/1925

llama.cpp v0.7.0 adds the LiliCorr speculative decoding technique and MTP for GLM5-Next, a new GPU cache for MoE experts kept in host memory (--moe-cache-mib, up to ~2.2x decode speedup when experts are offloaded), and a large batch of new models: K2 Horizon (dense and MoVA), Pri

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…

State-of-the-art probabilistic forecasting in four sizes (5M–209M params), CPU-efficient, zero-shot ready.

vllm-proto 0.6.0

📑 MinerU 4.x servers. The MinerU document engine now works with self-hosted MinerU 4.x servers, where every document failed with a 404 because 4.x removed the old parsing interface; 3.x servers work as before, 4.x servers take a tier (flash, basic, standard or advanced) and enab

🧱 External apps in channels. Administrators can now turn on Apps in Admin Settings > General, or with "ENABLE_APPS", and connect external apps in the new Apps tab of Admin Settings by their HTTPS address, reviewing the permissions each app asks for, such as reading its mentions,

All LiteLLM Docker images are signed with cosign. Every release is signed with the same key introduced in commit 0112e53.

Unsloth has gotten major performance upgrades!

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:

Models can now use the browser in Unsloth. Turn on Browser use and your model can open pages, read them, click, type and scroll while you watch. Qwen-Image-2.1-Turbo is here too, making images in just 8 steps. Also included: text-diffusion LLM training, faster MoE GGUFs and specu

All LiteLLM Docker images are signed with cosign. Every release is signed with the same key introduced in commit 0112e53.

What's Changed ollama launch claude no longer shows a warning about Claude connectors by @hoyyeva in #18878 ollama launch codex no longer warns about service tiers by @hoyyeva in #18879 Paused background model upgrade for embedding and other request performance, by @jmorganca…

☎️ Realtime voice calls. Administrators can now switch Call mode in the Voice calls section of Audio settings from Standard to Realtime, so Voice mode talks through an OpenAI Realtime voice model (gpt-realtime-2.1-mini with the Marin voice by default) that handles small talk itse

Release vllm-proto 0.5.0

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…

We're introducing Windows, Mac and Linux sandboxing in Unsloth! Read our guide. Turn any text or vision LLM into a Jev-style decision model in Unsloth, with decision accuracy going from 30% to 80%. Train, test, export and serve decision models directly from Unsloth. Also included

All LiteLLM Docker images are signed with cosign. Every release is signed with the same key introduced in commit 0112e53.

server: proxy cloud usage and balance APIs by @drifkin in https://github.com/ollama/ollama/pull/18829

Turn any text or vision LLM into a Jev-style decision model in Unsloth, with decision accuracy going from 30% to 80%. Train, test, export and serve decision models directly from Unsloth. Also included: native ComfyUI models, diffusion improvements and a better Browser in Desktop.

What's Changed 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. ollama pull qwen3.8 ollama run qwen3.8 Additional models include gemma4 , qwen3.6 and qwen3.5…

This release adds a browser inside Unsloth (browser use coming very soon), so files, web pages and pages the model writes open right beside your chat. It also brings EmbeddingGemma 2, Google's new multimodal embedding model, plus new Audio pages and lower-memory training.

llama.cpp v0.6.0 introduces the new llama_batch_ext extended batch API (with llama_process) for mixed token/embedding inputs and MTP/deepstack state embeddings, adds support for the GLM-5.3-Flash (GLM5-Next) 320B hybrid model, the Clef decision model (text and vision) and MTP spe

This release features 717 commits from 307 contributors (96 new)!

All LiteLLM Docker images are signed with cosign. Every release is signed with the same key introduced in commit 0112e53.

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…

779 PRs from 227 contributors.

This release brings faster navigation, shareable run settings, and clearer errors to Unsloth Desktop. It also 4x speeds up Laya decisions, expands hosted Decision API support, and keeps NVFP4, INT4, and MXFP4 checkpoints in 4-bit during LoRA training.

All LiteLLM Docker images are signed with cosign. Every release is signed with the same key introduced in commit 0112e53.

There has been 67 commits since v0.19.0 (September 10, 2026).

Release vllm-proto 0.4.0

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. Available…

Mistral opens a Munich hub for Physics AI and Industrial AI research, partnering with German industry.

We're adding support for decision models, a unified Library for docs and media, document viewer, many Apple Silicon improvements, creation of Skills, and ~4.5× faster image and video generation.

All LiteLLM Docker images are signed with cosign. Every release is signed with the same key introduced in commit 0112e53.

Prebuilt CUDA 13 wheels. Linux x86_64: flash-attn 2.8.4, built for PyTorch 2.13 and 2.14 on Python 3.13; causal-conv1d 1.7.0 and mamba-ssm 2.3.2.post1, built for PyTorch 2.10, 2.11 and 2.12 on Python 3.11, 3.12 and 3.13; 2.13 and 2.14 on Python 3.13. Windows x86_64: flash-attn 2.

All LiteLLM Docker images are signed with cosign. Every release is signed with the same key introduced in commit 0112e53.

This release focuses on backend performance and correctness, broader model coverage, and more robust server/router operation. It adds HRM-Text (DFM Mimir 1B) support, MiMo-V2.6 and HunyuanOCR conversion support, ggml 0.25.0 backend improvements, multi-address HTTP binding, image

You can now run Qwen-Image-2.1 locally with Unsloth! This release also includes custom Agent Skills, and easier chat/project management. It also brings 2x faster reasoning blocks (60 FPS vs 30 FPS), more reliable training, and improved Linux installs and updates. Qwen Image 2.1 G

All LiteLLM Docker images are signed with cosign. Every release is signed with the same key introduced in commit 0112e53.

Structured outputs on thinking models now apply in a single pass, making them faster and more reliable.

What's Changed 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"} '…

📉 Far smaller slim image. A slim build now comes down at around 175 MB, near enough 89% smaller than the last release, the local models, the packages around them and the tools that installed them all gone from it; what that changes about the way an instance behaves is set out un

v0.30.0 Highlights This release features 762 commits from 315 contributors (104 new)! New models : DeepSeek-V4.1-Flash ( #56214 , #56228 , #56208 ) with the whole KV stored in MXFP8 through the FlashMLA V4.1 record on SM100 ( #56893 ), DeepGEMM Mega-mHC ( #56962 ), and async…

713 PRs from 237 contributors.

We're releasing our new updated Docker image along with multi-user accounts, RDNA1+2, FP8/INT8 diffusion support, ARM64 CUDA Windows support and many training, GRPO and inference improvements. Also a Qwen3.8-Flash-Next 2x faster MTP hotfix.

Release vllm-proto 0.3.0

Open, private and multilingual AI is coming to your web browser. Mistral and Mozilla team up to put powerful, trustworthy AI where you already browse.

Validated by PR #56538 CI at fa2a26f .

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.

All LiteLLM Docker images are signed with cosign. Every release is signed with the same key introduced in commit 0112e53.

MLX safetensors ollama create no longer experimental. GGUF model creation now requires using llama.cpp tooling for safetensor conversion and quantization.

llama.cpp 0.4.1 adds Maple 20B-A1B, Tencent Hy 4, and Spark2.5 support, improves JSON schema handling, chat parsing, logging, and server child-process management, and updates ggml to v0.24.0.

vllm-proto 0.1.0

All LiteLLM Docker images are signed with cosign. Every release is signed with the same key introduced in commit 0112e53.

72 commits since v0.18.0 (July 17, 2026).

Cloudera and Mistral join forces to bring specialized, sovereign AI intelligence to enterprise data, helping regulated industries innovate on their own terms.

This is a large performance and reliability + bug fix release for Unsloth

Mistral helped a European energy operator migrate 40,000 lines of Fortran 77 to C++. Learn how it was done, and the lessons to carry forward.

This release features 594 commits from 277 contributors (91 new)!

Mistral today announced that it has raised €3 billion in a Series D funding round at a post-money valuation of more than €21 billion.

All LiteLLM Docker images are signed with cosign. Every release is signed with the same key introduced in commit 0112e53.

Ollama models can now be used directly in ChatGPT Desktop, so you can keep your existing workflow while running open models. Setup is available from the Ollama app on MacOS.

786 PRs from 214 contributors.

Run Qwen3.8-Flash-Next and GLM-5.3-Flash up to 2x faster with MTP. MTP is enabled by default, you can still disable it.

This release ships container images only. There is no PyPI package for `1.99.1`.

gemma4 now supports images and audio on MLX engine

A fix-focused release. The main fixes are for few-shot leakage, a multiple-choice filter bug, and group stderr, alongside two new ONNX backends and eight new benchmark suites. Also updated most configs for datasets>=4, which accounts for much of the diff by volume.

♿ Accessibility mode reaches the menus. Accessibility mode now marks the menu entry you are pointing at and the model already chosen with a stronger background, across the dropdown menus, their submenus, and the model picker together with its filter and compare controls, so those

🖼️ Richer previews for terminal files. Word documents and slide decks produced in the terminal are now previewed as the finished document rather than an approximation, and every document preview gains a page strip down the side with numbered thumbnails you can click to jump stra

Ollama's app now follows the system appearance again, restoring dark mode support

Qwen3.8-Flash-Next and GLM-5.3-Flash can now run locally in Unsloth!

MLX: Qwen3.8 Flash Next support

This release features 584 commits from 270 contributors (76 new)!

🚦 Human in the loop tool approval. Where an administrator has turned it on, you can switch a conversation from letting tools run freely to being asked first, so a model that wants to use a tool stops and waits for you to allow or deny it, one call at a time in a saved conversati

Thanks for the support for Qwen3.8-27B and Unsloth Desktop! This is a bug fix release with 170+ PRs.

710 PRs from 212 contributors.

Developers can now easily configure Claude Desktop to seamlessly work with Ollama as a third-party gateway provider.