### Run Reasonix Code

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/reasonix.md

Navigate to your project directory and execute this command to start Reasonix. No global installation is necessary. Reasonix defaults to DeepSeek-V4-Flash for cost-effective development.

```bash
cd /path/to/my-project
npx reasonix code
```

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

### Initialize and Run AstrBot

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/astrbot.md

Run these commands after installation. `astrbot init` should be run once to set up the environment in the current terminal directory. `astrbot run` starts the AstrBot service.

```bash
astrbot init
```

```bash
astrbot run
```

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

### Install AstrBot via Docker

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/astrbot.md

Clone the AstrBot repository and use Docker Compose to start the service. This method is suitable for users who prefer containerized deployments.

```bash
git clone https://github.com/AstrBotDevs/AstrBot --depth 1
cd AstrBot
```

```bash
sudo docker compose up -d
```

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

### Install Langcli (Manual)

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/langcli.md

Install Langcli globally using npm. Ensure Node.js 20 or later is installed.

```bash
npm i -g langcli-com
```

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

### Install Pi CLI

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/pi_mono.md

Install the Pi command-line interface globally using npm. Verify the installation by checking the version.

```bash
npm install -g @mariozechner/pi-coding-agent
```

```bash
pi --version
```

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

### Install nanobot using uv

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/nanobot.md

Use this command to install nanobot via the uv tool. Ensure uv is installed first.

```bash
uv tool install nanobot-ai
```

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

### Re-enter OpenClaw Configuration

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/openclaw.md

Run this command to re-enter the OpenClaw setup phase if it was previously installed. This is useful for changing configurations like the default model.

```bash
openclaw onboard --install-daemon
```

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

### Install and Launch OpenCode

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Download and install OpenCode, then launch it from your terminal. Connect to DeepSeek by selecting the provider in the TUI and entering your API key.

```bash
# Install from https://opencode.ai/download, then:
opencode

# In the TUI:
# Type /connect → enter "deepseek" → select the provider
# Enter your DeepSeek API Key
# Select DeepSeek-V4-Pro
```

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

### Install Kilo Code CLI

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/kilo_code.md

Install the Kilo Code CLI globally using npm. Ensure Node.js is installed first.

```bash
npm install -g @kilocode/cli
```

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

### Install DeepSeek CLI

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Install the DeepSeek CLI using npm or cargo. Ensure Rust 1.85+ is installed if building from source.

```bash
npm install -g deepseek-tui           # prebuilt cross-platform binary
```

```bash
cargo install deepseek-tui-cli        # build from source (Rust 1.85+)
```

```bash
deepseek --version
```

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

### Start nanobot agent

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/nanobot.md

Run this command in the terminal to start the nanobot agent after configuration.

```bash
nanobot agent
```

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

### Initialize and Run AstrBot

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Initialize the AstrBot environment once in your directory and then start the AstrBot service.

```bash
astrbot init   # run once to set up environment in current directory
```

```bash
astrbot run    # start AstrBot
```

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

### Install Crush CLI

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/crush.md

Install the Crush command-line interface globally using npm. Verify installation by checking the version.

```bash
npm install -g @charmland/crush
```

```bash
crush --version
```

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

### Start Langcli CLI

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/langcli.md

Run this command to start the Langcli interactive session. After starting, you can interact by typing commands like 'hi'.

```bash
langcli
```

```bash
# Then, in the session:
hi
```

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

### Install and Launch Kilo Code CLI

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Install Kilo Code using npm and launch it from your project directory. Connect to DeepSeek by selecting it in the TUI and entering your API key.

```bash
# Install (requires Node.js)
npm install -g @kilocode/cli
kilo --version

# Launch
cd /path/to/my-project
kilo

# In the TUI:
# Type /connect  → search "deepseek" → select DeepSeek → enter API key
# Type /models   → select one of:
#   DeepSeek Chat | DeepSeek Reasoner | DeepSeek V4 Flash | DeepSeek V4 Pro
```

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

### Install OpenClaw on Windows

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/openclaw.md

Use this command to install OpenClaw on Windows systems by downloading and executing the install script.

```powershell
iwr -useb https://openclaw.ai/install.ps1 | iex
```

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

### Install AstrBot via uv script

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Install AstrBot using a curl script for Linux/macOS or a PowerShell script for Windows. This is an alternative to Docker installation.

```bash
# Linux/macOS:
curl -LsSf https://docs.astrbot.app/install.sh | bash
```

```powershell
# Windows:
iwr -useb https://docs.astrbot.app/install.ps1 | iex
```

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

### Install OpenClaw on Linux/Mac

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/openclaw.md

Use this command to install OpenClaw on Linux or Mac systems by downloading and executing the install script.

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

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

### Install Pi via Official Script (Linux/macOS)

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/pi_mono.md

Alternative installation method for Linux and macOS users using a curl script.

```bash
curl -fsSL https://pi.dev/install.sh | sh
```

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

### Install Langcli (Windows)

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/langcli.md

Execute this command in an Administrator CMD on Windows to install Langcli.

```cmd
cmd /c "curl -fsSL -o %TEMP%\install-langcli.bat https://assets.langcli.com/installation/install-langcli.bat && %TEMP%\install-langcli.bat"
```

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

### Install and Run Crush

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Commands to install the Crush terminal harness and initiate its execution. After running, users can select the DeepSeek provider and model through an interactive menu.

```bash
cd /path/to/my-project
crush
# Press Ctrl+L or type /model → select "deepseek" provider → choose DeepSeek-V4-Pro
```

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

### Install AstrBot via uv (Windows)

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/astrbot.md

Use this command to install AstrBot on Windows systems. Ensure you have PowerShell.

```powershell
iwr -useb https://docs.astrbot.app/install.ps1 | iex
```

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

### Install Hermes Agent

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/hermes.md

Use this command to quickly install the Hermes Agent on Linux, macOS, or WSL2. Ensure Git is installed as a prerequisite.

```bash
curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash
```

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

### Install Pi Coding Agent

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Instructions for installing the Pi minimal, extensible terminal coding harness. It supports multiple providers and can be installed via npm or a shell script.

```bash
# Install
npm install -g @mariozechner/pi-coding-agent   # or: curl -fsSL https://pi.dev/install.sh | sh
pi --version
```

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

### Install AstrBot via uv (macOS/Linux)

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/astrbot.md

Use this command to install AstrBot on macOS and Linux systems. Ensure you have curl installed.

```bash
curl -LsSf https://docs.astrbot.app/install.sh | bash
```

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

### Verify nanobot installation

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/nanobot.md

Run this command to confirm nanobot has been installed successfully. A version number indicates success.

```bash
nanobot --version
```

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

### Install Langcli (macOS, Linux, WSL)

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/langcli.md

Use this command for a quick installation of Langcli on macOS, Linux, and WSL environments.

```bash
bash -c "$(curl -fsSL https://assets.langcli.com/installation/install-langcli.sh)"
```

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

### Verify DeepSeek-TUI Installation

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/deepseek-tui.md

Verify the installation of DeepSeek-TUI by checking its version. This command should be run after installation via npm or Cargo.

```sh
deepseek --version
```

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

### Install Codex CLI

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/codex.md

Install the Codex command-line interface globally using npm. Ensure Node.js 18+ is installed.

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

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

### Install GitHub Copilot CLI

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/copilot_cli.md

Install the GitHub Copilot CLI globally using npm. Requires Node.js 22 or later.

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

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

### nanobot Configuration Example

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/nanobot.md

This JSON structure defines default agent settings and provider details, including API keys and endpoints. Replace placeholders with your actual credentials.

```json
{
    "agents": {
        "defaults": {
            "model": "deepseek-v4-pro",
            "provider": "deepseek",
        }
    },
    "providers": {
        "deepseek": {
            "apiKey": "<your DeepSeek API Key>",
            "apiBase": "https://api.deepseek.com/v1",
        },
    },
}
```

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

### Verify Kilo Code CLI Installation

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/kilo_code.md

Check if the Kilo Code CLI is installed correctly by running the version command. A version number indicates successful installation.

```bash
kilo --version
```

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

### Install DeepSeek-TUI via npm

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/deepseek-tui.md

Install DeepSeek-TUI globally using npm for cross-platform prebuilt binaries. Ensure Node.js and npm are installed.

```sh
npm install -g deepseek-tui
```

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

### Verify Codex and Go Installation

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/codex.md

Check if the Codex CLI and Go are installed correctly by running their version commands.

```shell
codex --version
go version
```

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

### Start Moon Bridge Server

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/codex.md

Run the Moon Bridge application using the generated configuration file. This command starts the proxy server.

```shell
go run ./cmd/moonbridge --config config.yml
```

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

### Start Copilot CLI

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/copilot_cli.md

Launch the GitHub Copilot CLI in interactive mode. This command enables full agent mode, tool calling, and MCP support powered by DeepSeek.

```shell
copilot
```

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

### Install Claude Code CLI

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/claude_code.md

Install the Claude Code command-line interface globally using npm. Ensure Node.js 18+ is installed.

```bash
npm install -g @anthropic-ai/claude-code
```

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

### Install Deep Code CLI

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/deepcode.md

Install the Deep Code command-line interface globally using npm. Ensure Node.js version 18 or higher is installed.

```sh
npm install -g @vegamo/deepcode-cli
```

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

### Install Deep Code CLI

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Instructions for installing the Deep Code CLI, an AI coding assistant optimized for DeepSeek-V4. Requires Node.js 18+.

```bash
# Install (Node.js 18+)
npm install -g @vegamo/deepcode-cli
deepcode --version
```

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

### Launch Codex CLI

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/codex.md

Start the Codex CLI in your project directory. Ensure the `CODEX_HOME` environment variable is set to the directory where configuration files were generated.

```shell
cd /path/to/my-project
CODEX_HOME="$CODEX_HOME_DIR" codex --cd "$PWD"
```

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

### Install DeepSeek-TUI via Cargo

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/deepseek-tui.md

Install DeepSeek-TUI by building from source using Cargo. Rust 1.85 or later is required.

```sh
cargo install deepseek-tui-cli
```

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

### Launch DeepSeek-TUI in a Project Directory

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/deepseek-tui.md

Navigate to your project directory and launch DeepSeek-TUI. This command starts the interactive terminal AI coding assistant.

```sh
cd /path/to/my-project
deepseek
```

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

### Start Claude Code CLI Session

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/claude_code.md

Initiate a Claude Code CLI session by navigating to your project directory and running the 'claude' command.

```bash
cd /path/to/my-project
claude
```

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

### Configure and Run Moon Bridge

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Set up the Moon Bridge configuration file and start the server. Ensure the config.yml is correctly formatted with your DeepSeek API key and desired model settings.

```bash
git clone https://github.com/ZhiYi-R/moon-bridge.git && cd moon-bridge

cat > config.yml << 'EOF'
mode: "Transform"
server:
  addr: "127.0.0.1:38440"
provider:
  providers:
    deepseek:
      base_url: "https://api.deepseek.com/anthropic"
      api_key: "sk-your-deepseek-api-key"
      models:
        deepseek-v4-pro:
          context_window: 1000000
          max_output_tokens: 384000
          extensions:
            deepseek_v4:
              enabled: true
          default_reasoning_level: "high"
          supported_reasoning_levels:
            - effort: "high"
              description: "High reasoning effort"
            - effort: "xhigh"
              description: "Extra high reasoning effort"
          supports_reasoning_summaries: true
          default_reasoning_summary: "auto"
  routes:
    moonbridge:
      to: "deepseek/deepseek-v4-pro"
  default_model: "moonbridge"
EOF

# 3. Start Moon Bridge (keep terminal open)
go run ./cmd/moonbridge --config config.yml
```

```bash
# 4. Generate Codex config (in another terminal)
CODEX_HOME_DIR="${CODEX_HOME:-"$HOME/.codex"}"
mkdir -p "$CODEX_HOME_DIR"
MODEL="$(go run ./cmd/moonbridge --config config.yml --print-codex-model)"
go run ./cmd/moonbridge \
  --config config.yml \
  --print-codex-config "$MODEL" \
  --codex-base-url "http://127.0.0.1:38440/v1" \
  --codex-home "$CODEX_HOME_DIR" \
  > "$CODEX_HOME_DIR/config.toml"
```

```bash
# 5. Launch Codex
cd /path/to/my-project
CODEX_HOME="$CODEX_HOME_DIR" codex --cd "$PWD"
```

```bash
# One-command launcher alternative
./scripts/start_codex_with_moonbridge.sh --project-directory /path/to/my-project
```

```bash
# Verify
curl http://127.0.0.1:38440/v1/models
curl http://127.0.0.1:38440/v1/responses \
  -H "Content-Type: application/json" \
  -d '{"model":"moonbridge","input":"Say hello in one short sentence.","max_output_tokens":100}'
```

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

### Verify Deep Code Installation

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/deepcode.md

Check if the Deep Code CLI has been installed correctly by running the version command.

```sh
deepcode --version
```

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

### Install Codex Prerequisites

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Install the necessary prerequisites for using Codex with Moon Bridge, which translates requests to DeepSeek V4. This includes Node.js and Go.

```bash
# 1. Install prerequisites
npm install -g @openai/codex   # Node.js 18+
# Go 1.25+ also required
```

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

### Verify Claude Code CLI Installation

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/claude_code.md

Check if the Claude Code CLI was installed successfully by running the version command.

```bash
claude --version
```

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

### Run Deep Code CLI

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Commands to navigate to a project directory and start the Deep Code CLI. Includes a list of common key shortcuts for interacting with the TUI and information on agent skill locations.

```bash
cd /path/to/my-project
deepcode

# Key shortcuts:
#   Enter           → send prompt
#   Shift+Enter     → newline
#   Ctrl+V          → paste image from clipboard
#   Esc             → interrupt model turn
#   /               → open skills/commands menu
#   /new            → fresh conversation
#   /resume         → continue a previous conversation

# Agent Skills locations:
#   User-level:    ~/.agents/skills/<name>/SKILL.md
#   Project-level: ./.deepcode/skills/<name>/SKILL.md
```

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

### Configure Crush for DeepSeek

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Install Crush and set the DeepSeek API key as an environment variable. Configure the API key in the Crush configuration file located at ~/.config/crush/crush.json.

```bash
# Install
npm install -g @charmland/crush        # or: brew install charmbracelet/tap/crush
crush --version

# Set API key
export DEEPSEEK_API_KEY="sk-your-deepseek-api-key"   # Linux/macOS
# $env:DEEPSEEK_API_KEY="sk-your-deepseek-api-key"   # Windows PowerShell

# Configure: ~/.config/crush/crush.json  (Linux/macOS)
```

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

### Use DeepSeek V4 Flash with Oh My Pi

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/oh-my-pi.md

Start a project session with the DeepSeek V4 Flash model for faster responses. This requires the model to be defined in your `models.yml`.

```sh
omp --model deepseek/deepseek-v4-flash
```

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

### Configure Hermes Agent with DeepSeek

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Set up the Hermes Agent for DeepSeek by selecting 'Quick Setup', choosing DeepSeek as the provider, and entering your API key and preferred model.

```bash
hermes setup
# → Choose "Quick Setup"
# → Model provider: select DeepSeek
# → Enter your DeepSeek API Key
# → Base URL: https://api.deepseek.com
# → Model: deepseek-v4-pro
# → Continue with remaining options
```

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

### Configure GitHub Copilot CLI with DeepSeek

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Set up GitHub Copilot CLI to use DeepSeek models via the `anthropic` provider. This avoids the `reasoning_content` passback error. Environment variables are used for configuration. Ensure Node.js 22+ is installed.

```bash
# Install (requires Node.js 22+)
npm install -g @github/copilot

# Linux/macOS
export COPILOT_PROVIDER_TYPE=anthropic
export COPILOT_PROVIDER_BASE_URL=https://api.deepseek.com/anthropic
export COPILOT_PROVIDER_API_KEY=sk-your-deepseek-api-key
export COPILOT_MODEL=deepseek-v4-pro
# Optional: explicit token limits (needed because deepseek-v4-pro is not in built-in catalog)
export COPILOT_PROVIDER_MAX_PROMPT_TOKENS=840000
export COPILOT_PROVIDER_MAX_OUTPUT_TOKENS=128000
# Optional: block GitHub's telemetry calls while still routing prompts to DeepSeek
export COPILOT_OFFLINE=true

# Windows (PowerShell)
$env:COPILOT_PROVIDER_TYPE="anthropic"
$env:COPILOT_PROVIDER_BASE_URL="https://api.deepseek.com/anthropic"
$env:COPILOT_PROVIDER_API_KEY="sk-your-deepseek-api-key"
$env:COPILOT_MODEL="deepseek-v4-pro"

# Launch
copilot
# Available models: deepseek-v4-pro, deepseek-v4-flash
```

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

### Serve DeepSeek-TUI with HTTP Runtime API

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/deepseek-tui.md

Start DeepSeek-TUI as an HTTP server to expose its runtime API for integration with IDEs and web UIs. This allows programmatic access to sessions, threads, and tasks.

```sh
deepseek serve --http
```

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

### One-Command Codex Launcher Script

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/codex.md

Use the provided shell script to build and start the Moon Bridge proxy, generate Codex configuration, and launch Codex in a single command.

```shell
./scripts/start_codex_with_moonbridge.sh --project-directory /path/to/my-project
```

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

### Initialize nanobot configuration

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/nanobot.md

Execute this command to create the nanobot configuration file. The file path depends on your operating system.

```bash
nanobot onboard
```

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

### Run Pi and Select Model

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/pi_mono.md

Navigate to your project directory and run the `pi` command. Use the `/model` command within Pi to switch to the DeepSeek provider and select a specific model.

```bash
cd /path/to/my-project
pi
```

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

### Run Kilo Code in Project

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/kilo_code.md

Navigate to your project directory and execute the Kilo Code CLI. This command initiates the Kilo Code assistant within the project context.

```bash
cd /path/to/my-project
kilo
```

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

### Run Crush and Select Model

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/crush.md

Navigate to your project directory and run the 'crush' command. Use Ctrl+L to open the model switcher and select a DeepSeek model.

```bash
cd /path/to/my-project
crush
```

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

### Launch Deep Code in Project Directory

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/deepcode.md

Navigate to your project directory and launch the Deep Code assistant. This command initiates the AI coding assistant within the context of your project.

```sh
cd /path/to/my-project
deepcode
```

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

### Update shell for nanobot

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/nanobot.md

Update your current terminal session to recognize nanobot commands after installation or PATH changes.

```powershell
uv tool update-shell
```

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

### OpenAI-compatible Endpoint Configuration

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Configure the OpenAI-compatible endpoint for services like Crush or Pi. Set context_window and max_tokens in the provider configuration.

```bash
# Set context_window: 1000000 / max_tokens: 384000 in the provider config
BASE_URL=https://api.deepseek.com
MODEL=deepseek-v4-pro
```

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

### Launch OpenClaw Web UI

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/openclaw.md

Command to open the OpenClaw Web User Interface, allowing interaction through a browser.

```bash
openclaw dashboard
```

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

### Verify Moon Bridge Models Endpoint

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/codex.md

Check the available models exposed by the Moon Bridge server by sending a GET request to the `/v1/models` endpoint.

```shell
curl http://127.0.0.1:38440/v1/models
```

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

### Run Reasonix with npx

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Command to execute Reasonix, a DeepSeek-native terminal agent, using npx. This method does not require a global installation and is suitable for Node.js 20.10+.

```bash
# No global install needed — run via npx (Node.js 20.10+)
cd /path/to/my-project
npx reasonix code

# First run: built-in wizard prompts for your DeepSeek API key
# Config is persisted to ~/.reasonix/config.json

# In the TUI:
#   Default model: DeepSeek-V4-Flash (cost-efficient)
#   /pro          → switch to DeepSeek-V4-Pro for next turn
#   /preset max   → use V4-Pro for the whole session
#   /help         → full slash-command reference
```

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

### Add nanobot's bin directory to Windows PATH

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/nanobot.md

On Windows, add the .local\bin directory to your environment variables if nanobot commands are not recognized.

```powershell
$env:PATH = "$env:USERPROFILE\.local\bin;$env:PATH"
```

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

### Configure Logging Verbosity via Environment Variable

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/deepseek-tui.md

Set the logging verbosity for DeepSeek-TUI using the RUST_LOG environment variable. For example, set it to 'debug' for detailed logs.

```sh
export RUST_LOG=debug
```

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

### Use DeepSeek V4 Pro with Oh My Pi

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/oh-my-pi.md

Initiate a project session using the DeepSeek V4 Pro model. Ensure your `models.yml` is correctly configured.

```sh
cd /path/to/your-project
omp --model deepseek/deepseek-v4-pro
```

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

### DeepSeek CLI Authentication

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Set up authentication for the DeepSeek CLI. Use the interactive wizard or set the DEEPSEEK_API_KEY environment variable.

```bash
deepseek auth   # interactive wizard → saves to ~/.deepseek/config.toml
```

```bash
# Or set environment variable:
export DEEPSEEK_API_KEY="sk-your-deepseek-api-key"
```

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

### Configure Claude Code via Configuration File

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/claude_code.md

Set up Claude Code by creating a settings.json file with DeepSeek API credentials and model configurations. This file is used by both CLI and VSCode extension.

```json
{
  "env": {
    "ANTHROPIC_BASE_URL": "https://api.deepseek.com/anthropic",
    "ANTHROPIC_AUTH_TOKEN": "<your DeepSeek API Key>",
    "ANTHROPIC_MODEL": "deepseek-v4-pro[1m]",
    "ANTHROPIC_DEFAULT_OPUS_MODEL": "deepseek-v4-pro[1m]",
    "ANTHROPIC_DEFAULT_SONNET_MODEL": "deepseek-v4-pro[1m]",
    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "deepseek-v4-flash[1m]",
    "CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC": "1",
    "CLAUDE_CODE_EFFORT_LEVEL": "max"
  }
}
```

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

### Launch DeepSeek-TUI with a One-Shot Prompt

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/deepseek-tui.md

Launch DeepSeek-TUI and provide a prompt directly using the '-p' flag for a quick, one-off interaction.

```sh
deepseek -p "one-shot prompt"
```

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

### Configuring Reasoning Effort

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/CONTRIBUTING.md

Ensure guides are compatible with DeepSeek V4 Pro's 'max' reasoning effort level. For Claude Code, use 'CLAUDE_CODE_EFFORT_LEVEL=max'. For Anthropic-compatible endpoints, set 'thinking: { type: "enabled" }' with max budget. For OpenAI-compatible endpoints, include 'reasoning_effort' in the request body.

```text
CLAUDE_CODE_EFFORT_LEVEL=max
```

```yaml
thinking: {
  type: "enabled"
}
```

```json
{
  "reasoning_effort": "max"
}
```

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

### Configure Environment Variables for DeepSeek (Linux/Mac)

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/copilot_cli.md

Set environment variables to configure GitHub Copilot CLI to use DeepSeek models via the Anthropic provider. Ensure you replace 'sk-your-deepseek-api-key' with your actual DeepSeek API key.

```shell
export COPILOT_PROVIDER_TYPE=anthropic
export COPILOT_PROVIDER_BASE_URL=https://api.deepseek.com/anthropic
export COPILOT_PROVIDER_API_KEY=sk-your-deepseek-api-key
export COPILOT_MODEL=deepseek-v4-pro
```

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

### Pi Configuration for DeepSeek Models

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Configuration file for Pi, defining DeepSeek provider details, API type, models, context windows, and cost parameters. It includes compatibility settings for reasoning.

```json
{
  "providers": {
    "deepseek": {
      "baseUrl": "https://api.deepseek.com",
      "api": "openai-completions",
      "apiKey": "$DEEPSEEK_API_KEY",
      "models": [
        {
          "id": "deepseek-v4-pro",
          "name": "DeepSeek V4 Pro",
          "contextWindow": 1000000,
          "maxTokens": 384000,
          "input": ["text"],
          "reasoning": true,
          "cost": { "input": 1.74, "output": 3.48, "cacheRead": 0.145, "cacheWrite": 0 },
          "compat": {
            "requiresReasoningContentOnAssistantMessages": true,
            "thinkingFormat": "deepseek",
            "reasoningEffortMap": {
              "minimal": "high", "low": "high", "medium": "high",
              "high": "high", "xhigh": "max"
            }
          }
        },
        {
          "id": "deepseek-v4-flash",
          "name": "DeepSeek V4 Flash",
          "contextWindow": 1000000,
          "maxTokens": 384000,
          "input": ["text"],
          "reasoning": true,
          "cost": { "input": 0.14, "output": 0.28, "cacheRead": 0.028, "cacheWrite": 0 },
          "compat": {
            "requiresReasoningContentOnAssistantMessages": true,
            "thinkingFormat": "deepseek",
            "reasoningEffortMap": {
              "minimal": "high", "low": "high", "medium": "high",
              "high": "high", "xhigh": "max"
            }
          }
        }
      ]
    }
  }
}
```

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

### Configure DeepSeek Models for Oh My Pi

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/oh-my-pi.md

Create a `models.yml` file to define DeepSeek V4 models for Oh My Pi. This configuration includes essential compatibility settings for reliable tool usage.

```yaml
providers:
  deepseek:
    baseUrl: https://api.deepseek.com
    api: openai-completions
    apiKey: DEEPSEEK_API_KEY
    authHeader: true
    models:
      - id: deepseek-v4-pro
        name: DeepSeek V4 Pro
        reasoning: true
        thinking:
          minLevel: high
          maxLevel: xhigh
          mode: effort
        input: [text]
        contextWindow: 1000000
        maxTokens: 384000
        compat:
          supportsDeveloperRole: false
          supportsReasoningEffort: true
          maxTokensField: max_tokens
          reasoningEffortMap:
            high: high
            xhigh: max
          supportsToolChoice: false
          requiresReasoningContentForToolCalls: true
          requiresAssistantContentForToolCalls: true
          extraBody:
            thinking:
              type: enabled
      - id: deepseek-v4-flash
        name: DeepSeek V4 Flash
        reasoning: true
        thinking:
          minLevel: high
          maxLevel: xhigh
          mode: effort
        input: [text]
        contextWindow: 1000000
        maxTokens: 384000
        compat:
          supportsDeveloperRole: false
          supportsReasoningEffort: true
          maxTokensField: max_tokens
          reasoningEffortMap:
            high: high
            xhigh: max
          supportsToolChoice: false
          requiresReasoningContentForToolCalls: true
          requiresAssistantContentForToolCalls: true
          extraBody:
            thinking:
              type: enabled
```

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

### Configure Deep Code Settings

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/deepcode.md

Create a settings file to configure Deep Code with your DeepSeek API key, model, and other preferences. The `MODEL` option specifies the AI model to use, `BASE_URL` is the API endpoint, and `API_KEY` is your authentication token.

```json
{
  "env": {
    "MODEL": "deepseek-v4-pro",
    "BASE_URL": "https://api.deepseek.com",
    "API_KEY": "sk-..."
  },
  "thinkingEnabled": true,
  "reasoningEffort": "max"
}
```

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

### Configure GitHub Copilot Chat with DeepSeek V4

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Integrate DeepSeek V4 Pro & Flash into GitHub Copilot Chat via a VS Code extension. This preserves agent mode, tool calling, and skills. Ensure you have a GitHub Copilot subscription and set your DeepSeek API key.

```bash
# Prerequisites
# - VS Code 1.116+
# - GitHub Copilot subscription (free tier works)
# - Extension: https://marketplace.visualstudio.com/items?itemName=Vizards.deepseek-v4-for-copilot

# 1. Open Command Palette (Cmd+Shift+P / Ctrl+Shift+P)
#    Run: DeepSeek: Set API Key  →  paste your sk-ப்புகளை key

# 2. Open Copilot Chat (Cmd+Shift+I / Ctrl+Shift+I)
#    Click the model picker → choose "DeepSeek V4 Pro" or "DeepSeek V4 Flash"

# Optional: set thinking effort via the gear icon next to the model
#   None  | High (default) | Max

# Optional: vision proxy (DeepSeek V4 is text-only; extension auto-proxies images)
#   Command Palette → DeepSeek: Set Vision Proxy Model → pick Claude or GPT-4o
```

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

### DeepSeek-TUI Overview

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Introduction to DeepSeek-TUI, an open-source Rust terminal coding assistant. It supports the full 1M-token context window and sandboxed tool execution across multiple operating systems.

```bash

```

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

### Configure WorkBuddy/CodeBuddy for DeepSeek

Source: https://context7.com/deepseek-ai/awesome-deepseek-agent/llms.txt

Set the DeepSeek API key as an environment variable and create a models.json file to configure DeepSeek V4 models for WorkBuddy/CodeBuddy. Ensure the JSON file is saved as UTF-8 without BOM.

```powershell
# Set the API key as a persistent environment variable (Windows)
setx DEEPSEEK_API_KEY "sk-your-deepseek-api-key"
```

```json
# Create or edit: C:\Users\<username>\.codebuddy\models.json
# (or <your-project>\.codebuddy\models.json for project-scoped config)
# Save as UTF-8 WITHOUT BOM
{
  "models": [
    {
      "id": "deepseek-v4-pro",
      "name": "DeepSeek V4 Pro",
      "vendor": "DeepSeek",
      "url": "https://api.deepseek.com/v1/chat/completions",
      "apiKey": "${DEEPSEEK_API_KEY}",
      "maxInputTokens": 128000,
      "maxOutputTokens": 8192,
      "supportsToolCall": true,
      "supportsImages": false,
      "relatedModels": {
        "lite": "deepseek-v4-flash",
        "reasoning": "deepseek-v4-pro"
      }
    },
    {
      "id": "deepseek-v4-flash",
      "name": "DeepSeek V4 Flash",
      "vendor": "DeepSeek",
      "url": "https://api.deepseek.com/v1/chat/completions",
      "apiKey": "${DEEPSEEK_API_KEY}",
      "maxInputTokens": 128000,
      "maxOutputTokens": 8192,
      "supportsToolCall": true,
      "supportsImages": false
    }
  ],
  "availableModels": ["deepseek-v4-pro", "deepseek-v4-flash"]
}
```

```powershell
# Verify the API key before restarting the app
$env:DEEPSEEK_API_KEY="sk-your-deepseek-api-key"
curl https://api.deepseek.com/v1/chat/completions `
  -H "Content-Type: application/json" `
  -H "Authorization: Bearer $env:DEEPSEEK_API_KEY" `
  -d '{"model":"deepseek-v4-flash","messages":[{"role":"user","content":"hi"}],"stream":false}'
# Fully restart WorkBuddy/CodeBuddy, then select "DeepSeek V4 Pro" or "DeepSeek V4 Flash"
```

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

### Configure DeepSeek Provider for Crush

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/crush.md

Add DeepSeek AI models to your Crush configuration file. Ensure the API key is set as an environment variable.

```json
{
  "$schema": "https://charm.land/crush.json",
  "providers": {
    "deepseek": {
      "type": "openai-compat",
      "base_url": "https://api.deepseek.com",
      "api_key": "$DEEPSEEK_API_KEY",
      "models": [
        {
          "id": "deepseek-v4-pro",
          "name": "DeepSeek-V4-Pro",
          "context_window": 1048576,
          "default_max_tokens": 32768,
          "can_reason": true
        },
        {
          "id": "deepseek-v4-flash",
          "name": "DeepSeek-V4-Flash",
          "context_window": 1048576,
          "default_max_tokens": 32768,
          "can_reason": true
        }
      ]
    }
  }
}
```

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

### Launch DeepSeek-TUI in YOLO Mode

Source: https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/deepseek-tui.md

Launch DeepSeek-TUI in YOLO mode using the '--yolo' flag, which automatically approves all tool uses and lifts workspace boundaries.

```sh
deepseek --yolo
```