Get up and running with Kimi-K2.5, GLM-5, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models. https://ollama.com
Find a file
Jesse Gross ec3cc23076 mlxrunner: Release freed KV buffers during speculative decode
The decode loop releases MLX's pool of freed buffers every 256 generated
tokens, which is also how often the KV cache grows and drops its previous,
smaller buffers. The check fires only when the token count lands exactly on
a multiple of 256. Speculative decoding emits several tokens per round, so
most rounds step over the boundary and the pool is never released. Each
growth at a long context leaves several GB of buffers that no later
allocation can reuse, so the runner's footprint keeps climbing over a long
generation until the system runs out of memory.

We now release the pool whenever a round crosses a multiple of 256 tokens,
which is what a single-token round already did. With qwen3.8:27b-mlx at a
98k-token context on a 128 GB machine, a long speculative generation
previously grew the runner past 90 GB and panicked the kernel; it now stays
flat at 30 GB.
2026-09-17 09:45:01 -07:00
.github mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
anthropic Report cached prompt tokens (#17943) 2026-09-02 09:30:44 -07:00
api API: Deprecate typical_p (#18448) 2026-09-14 21:24:26 -07:00
app app: add Apps deep link (#18497) 2026-09-16 18:18:29 -07:00
auth auth: fix problems with the ollama keypairs (#12373) 2025-09-22 23:20:20 -07:00
cmake mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
cmd mlxrunner, cmd, server: fold three small packages into their users 2026-09-16 14:06:08 -07:00
create mlxrunner, cmd, server: fold three small packages into their users 2026-09-16 14:06:08 -07:00
discover Clean up dead code (#17381) 2026-08-27 14:33:51 -07:00
docs docs: fix cloud retirements settings link (#18496) 2026-09-16 16:20:58 -07:00
envconfig create: add server-side MLX imports and drop GGUF conversion (#14969) 2026-09-14 20:32:36 -07:00
format chore(all): replace instances of interface with any (#10067) 2025-04-02 09:44:27 -07:00
fs mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
harmony fix(docs): correct typos found during code review (#17579) 2026-08-31 08:59:55 -07:00
integration mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
internal create: add server-side MLX imports and drop GGUF conversion (#14969) 2026-09-14 20:32:36 -07:00
llama mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
llm create: add server-side MLX imports and drop GGUF conversion (#14969) 2026-09-14 20:32:36 -07:00
logutil runner: Remove CGO engines, use llama-server exclusively for GGML models (#16031) 2026-05-29 13:35:47 -07:00
manifest mlxrunner: read model manifests through the manifest package 2026-09-16 14:06:08 -07:00
middleware openai: finalize responses at the web search limit (#18328) 2026-09-08 16:27:20 -07:00
ml tokenizer, ml: remove dead code 2026-09-16 14:06:08 -07:00
mlx mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
mlxrunner mlxrunner: Release freed KV buffers during speculative decode 2026-09-17 09:45:01 -07:00
model mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
openai openai: support standalone named function outputs (#18348) 2026-09-09 14:51:50 -07:00
parser create: add server-side MLX imports and drop GGUF conversion (#14969) 2026-09-14 20:32:36 -07:00
progress tokenizer, ml: remove dead code 2026-09-16 14:06:08 -07:00
readline lint fixes (#17897) 2026-08-20 10:23:25 -07:00
scripts build: go deps (#18161) 2026-09-01 12:59:48 -07:00
server mlxrunner, cmd, server: fold three small packages into their users 2026-09-16 14:06:08 -07:00
template create: add server-side MLX imports and drop GGUF conversion (#14969) 2026-09-14 20:32:36 -07:00
thinking create: add server-side MLX imports and drop GGUF conversion (#14969) 2026-09-14 20:32:36 -07:00
tools tools: ignore braces inside JSON strings when detecting tool call end (#16937) 2026-06-27 12:00:55 -07:00
types API: Deprecate typical_p (#18448) 2026-09-14 21:24:26 -07:00
version add version 2023-08-22 09:40:58 -07:00
x/transfer mlxrunner: read model manifests through the manifest package 2026-09-16 14:06:08 -07:00
.dockerignore next build (#8539) 2025-01-29 15:03:38 -08:00
.gitattributes tokenizer, ml: remove dead code 2026-09-16 14:06:08 -07:00
.gitignore create: Clean up experimental paths, fix create from existing safetensor model (#14679) 2026-04-07 08:12:57 -07:00
.golangci.yaml mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
AGENTS.md Add AGENTS.md and CLAUDE.md to root repository (#16604) 2026-06-07 10:57:59 -07:00
CLAUDE.md Add AGENTS.md and CLAUDE.md to root repository (#16604) 2026-06-07 10:57:59 -07:00
CMakeLists.txt mlx: fix mac assumptions on linux/windows (#17898) 2026-08-20 09:50:10 -07:00
CMakePresets.json runner: Remove CGO engines, use llama-server exclusively for GGML models (#16031) 2026-05-29 13:35:47 -07:00
CONTRIBUTING.md docs: fix typos in repository documentation (#10683) 2025-11-15 20:22:29 -08:00
Dockerfile mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
go.mod tokenizer, ml: remove dead code 2026-09-16 14:06:08 -07:00
go.sum tokenizer, ml: remove dead code 2026-09-16 14:06:08 -07:00
LICENSE proto -> ollama 2023-06-26 15:57:13 -04:00
LLAMA_CPP_VERSION llama.cpp: version bump b10969 (#18446) 2026-09-15 13:13:31 -07:00
main.go lint 2024-08-01 17:06:06 -07:00
MLX_C_VERSION MLX, MLX-C: version bump (#18449) 2026-09-14 20:53:05 -07:00
MLX_VERSION MLX, MLX-C: version bump (#18449) 2026-09-14 20:53:05 -07:00
README.md launch: add DeepSeek Harness integration (#17733) 2026-08-13 15:19:02 -07:00
SECURITY.md docs: fix typos in repository documentation (#10683) 2025-11-15 20:22:29 -08:00

ollama

Ollama

Start building with open models.

Download

macOS

curl -fsSL https://ollama.com/install.sh | sh

or download manually

Windows

irm https://ollama.com/install.ps1 | iex

or download manually

Linux

curl -fsSL https://ollama.com/install.sh | sh

Manual install instructions

Docker

The official Ollama Docker image ollama/ollama is available on Docker Hub.

Libraries

Community

Get started

ollama

You'll be prompted to run a model or connect Ollama to your existing agents or applications such as Claude Code, OpenClaw, OpenCode , Codex, Copilot, and more.

Coding

To launch a specific integration:

ollama launch claude

Supported integrations include Claude Code, Codex, Copilot CLI, DeepSeek Harness, Droid, and OpenCode.

AI assistant

Use OpenClaw to turn Ollama into a personal AI assistant across WhatsApp, Telegram, Slack, Discord, and more:

ollama launch openclaw

Chat with a model

Run and chat with Gemma 4:

ollama run gemma4

See ollama.com/library for the full list.

See the quickstart guide for more details.

REST API

Ollama has a REST API for running and managing models.

curl http://localhost:11434/api/chat -d '{
  "model": "gemma4",
  "messages": [{
    "role": "user",
    "content": "Why is the sky blue?"
  }],
  "stream": false
}'

See the API documentation for all endpoints.

Python

pip install ollama
from ollama import chat

response = chat(model='gemma4', messages=[
  {
    'role': 'user',
    'content': 'Why is the sky blue?',
  },
])
print(response.message.content)

JavaScript

npm i ollama
import ollama from "ollama";

const response = await ollama.chat({
  model: "gemma4",
  messages: [{ role: "user", content: "Why is the sky blue?" }],
});
console.log(response.message.content);

Supported backends

  • llama.cpp project founded by Georgi Gerganov.

Documentation

Community Integrations

Want to add your project? Open a pull request.

Chat Interfaces

Web

Desktop

  • Dify.AI - LLM app development platform
  • AnythingLLM - All-in-one AI app for Mac, Windows, and Linux
  • Maid - Cross-platform mobile and desktop client
  • Witsy - AI desktop app for Mac, Windows, and Linux
  • Cherry Studio - Multi-provider desktop client
  • Ollama App - Multi-platform client for desktop and mobile
  • PyGPT - AI desktop assistant for Linux, Windows, and Mac
  • Alpaca - GTK4 client for Linux and macOS
  • SwiftChat - Cross-platform including iOS, Android, and Apple Vision Pro
  • Enchanted - Native macOS and iOS client
  • RWKV-Runner - Multi-model desktop runner
  • Ollama Grid Search - Evaluate and compare models
  • macai - macOS client for Ollama and ChatGPT
  • AI Studio - Multi-provider desktop IDE
  • Reins - Parameter tuning and reasoning model support
  • ConfiChat - Privacy-focused with optional encryption
  • LLocal.in - Electron desktop client
  • MindMac - AI chat client for Mac
  • Msty - Multi-model desktop client
  • BoltAI for Mac - AI chat client for Mac
  • IntelliBar - AI-powered assistant for macOS
  • Kerlig AI - AI writing assistant for macOS
  • Hillnote - Markdown-first AI workspace
  • Perfect Memory AI - Productivity AI personalized by screen and meeting history

Mobile

SwiftChat, Enchanted, Maid, Ollama App, Reins, and ConfiChat listed above also support mobile platforms.

Code Editors & Development

Libraries & SDKs

Frameworks & Agents

RAG & Knowledge Bases

  • RAGFlow - RAG engine based on deep document understanding
  • R2R - Open-source RAG engine
  • MaxKB - Ready-to-use RAG chatbot
  • Minima - On-premises or fully local RAG
  • Chipper - AI interface with Haystack RAG
  • ARGO - RAG and deep research on Mac/Windows/Linux
  • Archyve - RAG-enabling document library
  • Casibase - AI knowledge base with RAG and SSO
  • BrainSoup - Native client with RAG and multi-agent automation

Bots & Messaging

Terminal & CLI

Productivity & Apps

Observability & Monitoring

  • Opik - Debug, evaluate, and monitor LLM applications
  • OpenLIT - OpenTelemetry-native monitoring for Ollama and GPUs
  • Lunary - LLM observability with analytics and PII masking
  • Langfuse - Open source LLM observability
  • HoneyHive - AI observability and evaluation for agents
  • MLflow Tracing - Open source LLM observability

Database & Embeddings

Infrastructure & Deployment

Cloud

Package Managers