Ollama

https://docs.ollama.com/
Admin
Ollama is the easiest way to get up and running with large language models such as gpt-oss, Gemma 3,...
Tokens:91,739
Snippets:894
Trust Score:9.9
Update:2 weeks ago
Show doc for...
Tokens:
Raw
### Quick setup: launch Pi with Ollama

Source: https://docs.ollama.com/integrations/pi

Installs Pi if needed, configures Ollama as a provider including web tools, and starts an interactive session.

```bash
ollama launch pi

```

--------------------------------

### Start Ollama server

Source: https://docs.ollama.com/linux

Starts the Ollama server. Run this in a terminal after installation.

```bash
ollama serve

```

--------------------------------

### Quick setup with ollama launch omp

Source: https://docs.ollama.com/integrations/oh-my-pi

Quick setup command that configures Ollama as the model provider, sets up web search tools, and starts an interactive OMP session.

```bash
ollama launch omp

```

--------------------------------

### Launch Hermes Desktop with Ollama

Source: https://docs.ollama.com/integrations/hermes-desktop

Use this command to start Hermes Desktop. Ollama handles the setup flow automatically: installing the Hermes command-line agent if needed, letting you pick a model, configuring Hermes Desktop to use the selected model, and launching the app.

```bash
ollama launch hermes-desktop
```

--------------------------------

### Quick setup: ollama launch pool

Source: https://docs.ollama.com/integrations/pool

Installs and launches Pool with Ollama using the default configuration.

```bash
ollama launch pool

```

--------------------------------

### Install Hermes directly

Source: https://docs.ollama.com/integrations/hermes

Installs Hermes directly, bypassing `ollama launch hermes`, and launches the setup wizard automatically.

```bash
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
```

--------------------------------

### Install the Ollama JavaScript library (npm/bun)

Source: https://docs.ollama.com/capabilities/tool-calling

Installs the Ollama JavaScript library using npm or bun. Run either command to set up the dependency before using the JavaScript example.

```bash
# with npm
npm i ollama

# with bun
bun i ollama
```

--------------------------------

### Hermes setup wizard prompt

Source: https://docs.ollama.com/integrations/hermes

Shows the setup wizard's initial prompt; choose Quick setup for provider, model, and messaging.

```text
How would you like to set up Hermes?

 →  Quick setup — provider, model & messaging (recommended)
    Full setup — configure everything
```

--------------------------------

### Configure Pi without launching

Source: https://docs.ollama.com/integrations/pi

Runs the same setup as the quick launch but only configures Ollama as a provider without starting an interactive session.

```bash
ollama launch pi --config

```

--------------------------------

### Install the Ollama Python SDK (pip/uv)

Source: https://docs.ollama.com/capabilities/tool-calling

Installs the Ollama Python SDK using pip or uv. Run these commands in your terminal before using the Python examples on this page.

```bash
# with pip
pip install ollama -U

# with uv
uv add ollama    
```

--------------------------------

### Launch integrations interactively with ollama launch

Source: https://docs.ollama.com/cli

Launches the interactive integration setup wizard. Use 'ollama launch' to configure and start external applications that use Ollama models.

```bash
ollama launch
```

--------------------------------

### Example /loop commands

Source: https://docs.ollama.com/integrations/claude-code

Example /loop commands with different intervals and prompts.

```bash
/loop 30m Check my open PRs and summarize their status
```

```bash
/loop 1h Research the latest AI news and summarize key developments
```

```bash
/loop 15m Check for new GitHub issues and triage by priority
```

--------------------------------

### Quick setup and run with a model

Source: https://docs.ollama.com/integrations/copilot-cli

Quick setup launches Copilot CLI with Ollama. The second command runs directly with a specified model, here kimi-k2.5:cloud.

```bash
ollama launch copilot
```

```bash
ollama launch copilot --model kimi-k2.5:cloud
```

--------------------------------

### Start Ollama server with ollama serve

Source: https://docs.ollama.com/cli

Starts the Ollama server. Use 'ollama serve --help' to view environment variables that can be set.

```bash
ollama serve
```

--------------------------------

### Manual setup with ollama launch omp --config

Source: https://docs.ollama.com/integrations/oh-my-pi

Manual setup command to configure OMP after installing it from omp.sh. The Ollama web search plugin is managed automatically when launching through Ollama.

```bash
ollama launch omp --config

```

--------------------------------

### Launch Hermes with Ollama

Source: https://docs.ollama.com/integrations/hermes

Starts the Hermes Agent with Ollama handling installation, model selection, onboarding, and optional messaging gateway setup.

```bash
ollama launch hermes
```

--------------------------------

### Manual setup with environment variables

Source: https://docs.ollama.com/integrations/copilot-cli

Manual setup: set environment variables to connect Copilot CLI to Ollama's OpenAI-compatible API, then run copilot. Alternatively, run with environment variables inline. Copilot requires a large context window (at least 64k tokens).

```bash
export COPILOT_PROVIDER_BASE_URL=http://localhost:11434/v1
export COPILOT_PROVIDER_API_KEY=
export COPILOT_PROVIDER_WIRE_API=responses
export COPILOT_MODEL=qwen3.5
```

```bash
copilot
```

```bash
COPILOT_PROVIDER_BASE_URL=http://localhost:11434/v1 COPILOT_PROVIDER_API_KEY= COPILOT_PROVIDER_WIRE_API=responses COPILOT_MODEL=glm-5:cloud copilot
```

--------------------------------

### Install Cline CLI with npm

Source: https://docs.ollama.com/integrations/cline-cli

Installs the Cline CLI globally. If Cline CLI is not installed and npm is available, `ollama launch cline` will prompt to install `cline@latest`.

```bash
npm install -g cline

```

--------------------------------

### Install Pi extensions from npm or git

Source: https://docs.ollama.com/integrations/pi

Installs on-demand capability packages via the /skill:name command. Packages can be installed from npm or git repositories.

```bash
pi install npm:@foo/some-tools
pi install git:github.com/user/repo@v1

```

--------------------------------

### Install OpenCode on Windows

Source: https://docs.ollama.com/integrations/opencode

Installs OpenCode globally on Windows via npm.

```bash
npm install -g opencode-ai
```

--------------------------------

### Install to a standard prefix

Source: https://docs.ollama.com/development

Installs the built Ollama into a specified prefix directory using CMake's install command.

```bash
cmake --install build --prefix /path/to/install

```

--------------------------------

### Verify Ollama installation

Source: https://docs.ollama.com/linux

Verifies that Ollama is running by printing its version. Run in another terminal after starting the server.

```bash
ollama -v

```

--------------------------------

### Pull a model for manual setup

Source: https://docs.ollama.com/integrations/cline-cli

Pulls the `qwen3.5` model to make it available for manual setup. Ensure Ollama is running and the model is available before configuring Cline CLI manually.

```bash
ollama pull qwen3.5

```

--------------------------------

### Install OpenCode on macOS/Linux

Source: https://docs.ollama.com/integrations/opencode

Installs OpenCode on macOS or Linux using the official installation script.

```bash
curl -fsSL https://opencode.ai/install | bash
```

--------------------------------

### Install NemoClaw with Ollama provider

Source: https://docs.ollama.com/integrations/nemoclaw

Runs the NemoClaw installer with Ollama as the provider and the recommended model. Requires Ollama to be installed and running before execution.

```bash
curl -fsSL https://www.nvidia.com/nemoclaw.sh | \
  NEMOCLAW_NON_INTERACTIVE=1 \
  NEMOCLAW_PROVIDER=ollama \
  NEMOCLAW_MODEL=nemotron-3-nano:30b \
  bash
```

--------------------------------

### Install marimo with uvx sandbox

Source: https://docs.ollama.com/integrations/marimo

Run this command to install and launch marimo in a sandboxed environment using uv. The notebook file is named notebook.py.

```bash
uvx marimo edit --sandbox notebook.py

```

--------------------------------

### Install NVIDIA Container Toolkit with Apt

Source: https://docs.ollama.com/docker

Installs the NVIDIA Container Toolkit packages on Debian/Ubuntu systems.

```bash
sudo apt-get install -y nvidia-container-toolkit

```

--------------------------------

### Install Pi globally via npm

Source: https://docs.ollama.com/integrations/pi

Installs the Pi coding agent globally using npm.

```bash
npm install -g @earendil-works/pi-coding-agent

```

--------------------------------

### Install web search package manually

Source: https://docs.ollama.com/integrations/pi

Installs the @ollama/pi-web-search package manually. When launching Pi through Ollama, package install/update is managed automatically.

```bash
pi install npm:@ollama/pi-web-search

```

--------------------------------

### Install pi-autoresearch extension

Source: https://docs.ollama.com/integrations/pi

Installs the pi-autoresearch package from GitHub, which brings autonomous experiment loops to Pi.

```bash
pi install https://github.com/davebcn87/pi-autoresearch

```

--------------------------------

### Start and verify Ollama service

Source: https://docs.ollama.com/linux

Starts the Ollama service and checks its status. Use after enabling the service.

```bash
sudo systemctl start ollama
sudo systemctl status ollama

```

--------------------------------

### Install Copilot CLI

Source: https://docs.ollama.com/integrations/copilot-cli

Install GitHub Copilot CLI on macOS/Linux via Homebrew, on all platforms via npm, on macOS/Linux via the official script, or on Windows via WinGet.

```bash
brew install copilot-cli
```

```bash
npm install -g @github/copilot
```

```bash
curl -fsSL https://gh.io/copilot-install | bash
```

```bash
winget install GitHub.Copilot
```

--------------------------------

### Search agent example output

Source: https://docs.ollama.com/capabilities/web-search

Example output from the search agent showing the model's thinking, tool calls, and final content response. It demonstrates how the agent uses web_search to gather information and then synthesizes a response.

```text
Thinking:  Okay, the user is asking about Ollama's new engine. I need to figure out what they're referring to. Ollama is a company that develops large language models, so maybe they've released a new model or an updated version of their existing engine....

Tool calls:  [ToolCall(function=Function(name='web_search', arguments={'max_results': 3, 'query': 'Ollama new engine'}))]
Result:  results=[WebSearchResult(content='# New model scheduling\n\n## September 23, 2025\n\nOllama now includes a significantly improved model scheduling system. Ahead of running a model, Ollama’s new engine

Thinking:  Okay, the user asked about Ollama's new engine. Let me look at the search results.

First result is from September 23, 2025, talking about new model scheduling. It mentions improved memory management, reduced crashes, better GPU utilization, and multi-GPU performance. Examples show speed improvements and accurate memory reporting. Supported models include gemma3, llama4, qwen3, etc...

Content:  Ollama has introduced two key updates to its engine, both released in 2025:

1. **Enhanced Model Scheduling (September 23, 2025)**
   - **Precision Memory Management**: Exact memory allocation reduces out-of-memory crashes and optimizes GPU utilization.
   - **Performance Gains**: Examples show significant speed improvements (e.g., 85.54 tokens/s vs 52.02 tokens/s) and full GPU layer utilization.
   - **Multi-GPU Support**: Improved efficiency across multiple GPUs, with accurate memory reporting via tools like `nvidia-smi`.
   - **Supported Models**: Includes `gemma3`, `llama4`, `qwen3`, `mistral-small3.2`, and more.

2. **Multimodal Engine (May 15, 2025)**
   - **Vision Support**: First-class support for vision models, including `llama4:scout` (109B parameters), `gemma3`, `qwen2.5vl`, and `mistral-small3.1`.
   - **Multimodal Tasks**: Examples include identifying animals in multiple images, answering location-based questions from videos, and document scanning.

These updates highlight Ollama's focus on efficiency, performance, and expanded capabilities for both text and vision tasks.

```

--------------------------------

### Set environment variables for manual setup

Source: https://docs.ollama.com/integrations/claude-code

Sets environment variables for manual setup to connect Claude Code to a local Ollama instance.

```bash
export ANTHROPIC_AUTH_TOKEN=ollama
export ANTHROPIC_API_KEY=""
export ANTHROPIC_BASE_URL=http://localhost:11434
```

--------------------------------

### Run OpenCode

Source: https://docs.ollama.com/integrations/opencode

Runs OpenCode after configuration, starting an interactive session.

```bash
opencode
```

--------------------------------

### Example response for show model details

Source: https://docs.ollama.com/api-reference/show-model-details

Example 200 response from the /api/show endpoint for the gemma4 model. Includes parameters, license, capabilities, thinking settings, and detailed model_info.

```json
{
  "parameters": "temperature 0.7\nnum_ctx 2048",
  "license": "Gemma Terms of Use \n\nLast modified: February 21, 2024...",
  "capabilities": [
    "completion",
    "thinking",
    "vision"
  ],
  "thinking": {
    "values": [
      false,
      true
    ],
    "default": true
  },
  "modified_at": "2025-08-14T15:49:43.634137516-07:00",
  "details": {
    "parent_model": "",
    "format": "gguf",
    "family": "gemma4",
    "families": [
      "gemma4"
    ],
    "parameter_size": "8.0B",
    "quantization_level": "Q4_K_M"
  },
  "model_info": {
    "gemma4.attention.head_count": 8,
    "gemma4.attention.head_count_kv": 2,
    "gemma4.attention.key_length": 512,
    "gemma4.attention.key_length_swa": 256,
    "gemma4.attention.layer_norm_rms_epsilon": 0.000001,
    "gemma4.attention.shared_kv_layers": 18,
    "gemma4.attention.sliding_window": 512,
    "gemma4.attention.value_length": 512,
    "gemma4.attention.value_length_swa": 256,
    "gemma4.audio.attention.head_count": 8,
    "gemma4.audio.attention.layer_norm_epsilon": 0.000001,
    "gemma4.audio.block_count": 12,
    "gemma4.audio.conv_kernel_size": 5,
    "gemma4.audio.embedding_length": 1024,
    "gemma4.audio.feed_forward_length": 4096,
    "gemma4.block_count": 42,
    "gemma4.context_length": 131072,
    "gemma4.embedding_length": 2560,
    "gemma4.embedding_length_per_layer_input": 256,
    "gemma4.feed_forward_length": 10240,
    "gemma4.final_logit_softcapping": 30,
    "gemma4.rope.dimension_count": 512,
    "gemma4.rope.dimension_count_swa": 256,
    "gemma4.rope.freq_base": 1000000,
    "gemma4.rope.freq_base_swa": 10000,
    "gemma4.vision.attention.head_count": 12,
    "gemma4.vision.attention.layer_norm_epsilon": 0.000001,
    "gemma4.vision.block_count": 16,
    "gemma4.vision.embedding_length": 768,
    "gemma4.vision.feed_forward_length": 3072,
    "gemma4.vision.num_channels": 3,
    "gemma4.vision.patch_size": 16,
    "gemma4.vision.projector.scale_factor": 3,
    "general.architecture": "gemma4",
    "general.file_type": 15,
    "general.quantization_version": 2,
    "tokenizer.ggml.add_bos_token": false,
    "tokenizer.ggml.add_eos_token": false,
    "tokenizer.ggml.add_mask_token": false,
    "tokenizer.ggml.add_padding_token": false,
    "tokenizer.ggml.add_unknown_token": false,
    "tokenizer.ggml.bos_token_id": 2,
    "tokenizer.ggml.eos_token_id": 1,
    "tokenizer.ggml.eos_token_ids": [
      1,
      106,
      50
    ],
    "tokenizer.ggml.mask_token_id": 4,
    "tokenizer.ggml.merges": null,
    "tokenizer.ggml.model": "llama",
    "tokenizer.ggml.padding_token_id": 0,
    "tokenizer.ggml.pre": "gemma4",
    "tokenizer.ggml.scores": null,
    "tokenizer.ggml.token_type": null,
    "tokenizer.ggml.tokens": null,
    "tokenizer.ggml.unknown_token_id": 3
  }
}
```

--------------------------------

### Profile-based setup: create ~/.codex/ollama-launch.config.toml

Source: https://docs.ollama.com/integrations/codex

Create a persistent Codex CLI configuration file at `~/.codex/ollama-launch.config.toml` with the specified content. This configures the Ollama provider and model catalog.

```toml
model = "gpt-oss:120b"
model_provider = "ollama-launch"
model_catalog_json = "/Users/you/.codex/model.json"

[model_providers.ollama-launch]
name = "Ollama"
base_url = "http://localhost:11434/v1/"
wire_api = "responses"

```

--------------------------------

### 200 response for GET /api/ps

Source: https://docs.ollama.com/api/ps

Example JSON response from the GET /api/ps endpoint. The 'models' array contains objects with details such as name, size, digest, and context_length for each running model.

```json
{
  "models": [
    {
      "name": "gemma4",
      "model": "gemma4",
      "size": 6591830464,
      "digest": "c6eb396dbd5992bbe3f5cdb947e8bbc0ee413d7c17e2beaae69f5d569cf982eb",
      "details": {
        "parent_model": "",
        "format": "gguf",
        "family": "gemma4",
        "families": [
          "gemma4"
        ],
        "parameter_size": "8.0B",
        "quantization_level": "Q4_K_M"
      },
      "expires_at": "2025-10-17T16:47:07.93355-07:00",
      "size_vram": 5333539264,
      "context_length": 4096
    }
  ]
}
```

--------------------------------

### Manual setup with --oss flag

Source: https://docs.ollama.com/integrations/codex

Use the `--oss` flag to run Codex with Ollama.

```bash
codex --oss

```

--------------------------------

### Run Ollama from repository root

Source: https://docs.ollama.com/development

Starts the Ollama server using the Go binary from the repository root. Requires Go and a native payload already built.

```bash
go run . serve

```

--------------------------------

### Response for GET /api/version

Source: https://docs.ollama.com/api-reference/get-version

Example JSON response from the version endpoint, showing the version field with the value "0.12.6".

```json
{
  "version": "0.12.6"
}
```

--------------------------------

### List models response (200)

Source: https://docs.ollama.com/api/tags

Example JSON response from the GET /api/tags endpoint, showing a model named 'gemma4' with its metadata.

```json
{
  "models": [
    {
      "name": "gemma4",
      "model": "gemma4",
      "modified_at": "2025-10-03T23:34:03.409490317-07:00",
      "size": 9608350245,
      "digest": "c6eb396dbd5992bbe3f5cdb947e8bbc0ee413d7c17e2beaae69f5d569cf982eb",
      "details": {
        "format": "gguf",
        "family": "gemma4",
        "families": [
          "gemma4"
        ],
        "parameter_size": "8.0B",
        "quantization_level": "Q4_K_M"
      }
    }
  ]
}
```

--------------------------------

### MESSAGE example conversation

Source: https://docs.ollama.com/modelfile

Use multiple iterations of the MESSAGE command to build up a conversation which will guide the model to answer in a similar way.

```text
MESSAGE user Is Toronto in Canada?
MESSAGE assistant yes
MESSAGE user Is Sacramento in Canada?
MESSAGE assistant no
MESSAGE user Is Ontario in Canada?
MESSAGE assistant yes
```

--------------------------------

### Basic Modelfile example (mario blueprint)

Source: https://docs.ollama.com/modelfile

Creates a custom model based on llama3.2 with temperature 1, context window 4096, and a custom system message. Save as a file and use with 'ollama create' and 'ollama run'.

```modelfile
FROM llama3.2
# sets the temperature to 1 [higher is more creative, lower is more coherent]
PARAMETER temperature 1
# sets the context window size to 4096, this controls how many tokens the LLM can use as context to generate the next token
PARAMETER num_ctx 4096

# sets a custom system message to specify the behavior of the chat assistant
SYSTEM You are Mario from super mario bros, acting as an assistant.

```

--------------------------------

### Native build and run

Source: https://docs.ollama.com/development

Builds the native runtime and Go binary from the repository root, then runs the server. On macOS arm64 this builds Metal inference; on other platforms CPU-only.

```bash
cmake -B build .
cmake --build build --parallel 8
./ollama serve

```

--------------------------------

### Install specific Ollama version on Linux

Source: https://docs.ollama.com/troubleshooting

Installs a specific version of Ollama on Linux by setting the OLLAMA_VERSION environment variable when running the install script. Useful for installing older or pre-release versions.

```bash
curl -fsSL https://ollama.com/install.sh | OLLAMA_VERSION=0.5.7 sh

```

--------------------------------

### Quick setup with ollama launch cline

Source: https://docs.ollama.com/integrations/cline-cli

Launches Cline CLI through Ollama. When launched this way, Ollama sets Cline's provider to Ollama, points it at the local Ollama endpoint, and selects the model you choose.

```bash
ollama launch cline

```

--------------------------------

### Create a customized model with Modelfile

Source: https://docs.ollama.com/cli

Shows how to create a custom model by defining a Modelfile and then running 'ollama create'. The Modelfile specifies the base model and system prompt.

```dockerfile
FROM gemma4
SYSTEM """You are a happy cat."""
```

```bash
ollama create -f Modelfile
```

--------------------------------

### Run Ollama CLI

Source: https://docs.ollama.com/cloud

Launch the Ollama command-line interface. After running 'ollama', follow the setup prompts and sign in to use cloud models.

```bash
ollama

```

--------------------------------

### Manually install DeepSeek Harness

Source: https://docs.ollama.com/integrations/deepseek-harness

Manually installs the latest DeepSeek Harness globally. Requires Node.js; on Windows, install Node.js for Windows first.

```shell
npm install -g @deepseek-ai/dsh@latest
```

--------------------------------

### Connect messaging platform prompt

Source: https://docs.ollama.com/integrations/hermes

Optional step to connect a messaging platform during setup; choose to set up now or skip for later.

```text
Connect a messaging platform? (Telegram, Discord, etc.)

 →  Set up messaging now (recommended)
    Skip — set up later with 'hermes gateway setup'
```

--------------------------------

### Install Metal Toolchain on macOS

Source: https://docs.ollama.com/development

Install the Metal toolchain required for MLX Metal on macOS (Apple Silicon). Run this command after installing Xcode.

```bash
xcodebuild -downloadComponent MetalToolchain
```

--------------------------------

### Start interactive Cline session

Source: https://docs.ollama.com/integrations/cline-cli

Starts an interactive Cline session.

```bash
cline

```

--------------------------------

### TEMPLATE instruction example

Source: https://docs.ollama.com/modelfile

Defines the full prompt template passed to the model. Uses Go template syntax and may include system, user, and response variables. The response variable is omitted when generating a response.

```modelfile
TEMPLATE """{{ if .System }}<|im_start|>system
{{ .System }}<|im_end|>
{{ end }}{{ if .Prompt }}<|im_start|>user
{{ .Prompt }}<|im_end|>
{{ end }}<|im_start|>assistant
"""

```

--------------------------------

### Launch OpenCode with Ollama

Source: https://docs.ollama.com/cloud

Launch OpenCode as a coding agent using Ollama's cloud. Requires the Ollama app or CLI to be set up and signed in.

```bash
ollama launch opencode

```

--------------------------------

### Install Claude Code

Source: https://docs.ollama.com/integrations/claude-code

Install Claude Code on macOS/Linux (bash) or Windows (PowerShell).

```bash
curl -fsSL https://claude.ai/install.sh | bash
```

```powershell
irm https://claude.ai/install.ps1 | iex
```

--------------------------------

### Launch OpenCode with Ollama

Source: https://docs.ollama.com/integrations/opencode

Launches OpenCode with Ollama as the provider, using an inline configuration for the selected Ollama model.

```bash
ollama launch opencode
```

--------------------------------

### Install Ollama to a custom directory with /DIR flag

Source: https://docs.ollama.com/windows

Use this command to install the Ollama application to a custom directory instead of the default home directory. The /DIR flag specifies the target installation path.

```powershell
OllamaSetup.exe /DIR="d:\some\location"

```

--------------------------------

### Manual Setup Configuration

Source: https://docs.ollama.com/integrations/droid

Add a local configuration block to ~/.factory/config.json to manually set up Droid with Ollama. Uses the qwen3-coder model with a local base URL and max_tokens of 32000.

```json
{
  "custom_models": [
    {
      "model_display_name": "qwen3-coder [Ollama]",
      "model": "qwen3-coder",
      "base_url": "http://localhost:11434/v1/",
      "api_key": "not-needed",
      "provider": "generic-chat-completion-api",
      "max_tokens": 32000 
    }
  ]
}

```

--------------------------------

### Configure NVIDIA Container Toolkit repository with Apt

Source: https://docs.ollama.com/docker

Configures the NVIDIA Container Toolkit repository on Debian/Ubuntu systems using apt.

```bash
curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey \
    | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg
curl -fsSL https://nvidia.github.io/libnvidia-container/stable/deb/nvidia-container-toolkit.list \
    | sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' \
    | sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list
sudo apt-get update

```

--------------------------------

### Prompt coding agent to explain repository

Source: https://docs.ollama.com/quickstart

Example prompt to use with a coding agent after choosing a model. This is a sample request to understand repository structure.

```text
Explain how this repository is organized.

```

--------------------------------

### Install Codex CLI

Source: https://docs.ollama.com/integrations/codex

Install the Codex CLI globally using npm. This is required before using Codex with Ollama.

```bash
npm install -g @openai/codex

```

--------------------------------

### Configure DeepSeek Harness without starting

Source: https://docs.ollama.com/integrations/deepseek-harness

Opens the configuration interface for DeepSeek Harness without starting the agent.

```shell
ollama launch dsh --config
```
Tokens:
Raw