Pydantic AI

https://pydantic.dev/docs/ai/
Admin
Pydantic AI is a Python agent framework designed to build production-grade applications and...
Tokens:2,346,882
Snippets:15,864
Trust Score:9.5
Update:2 days ago
Show doc for...
Tokens:
Raw
### Install dependencies and run coding agent example

Source: https://pydantic.dev/docs/ai/harness/examples/

Commands to install dependencies and run the coding_agent.py example from the repo root. Requires make and uv installed.

```bash
make install
uv run examples/coding_agent.py
```

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

### Run the example with uv

Source: https://pydantic.dev/docs/ai/examples/getting-started/pydantic-model/

Run the example using uv after dependencies are installed and environment variables are set.

```bash
uv run -m pydantic_ai_examples.pydantic_model
```

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

### Install uvicorn using uv

Source: https://pydantic.dev/docs/ai/integrations/ui/ag-ui/

Install uvicorn, an ASGI server, to run the examples.

```bash
uv add uvicorn
```

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

### Run a Pydantic AI example with uv one-liner and zero setup

Source: https://pydantic.dev/docs/ai/examples/setup/

Run a pydantic-ai example with zero setup using uv, setting the OPENAI_API_KEY inline and using --with to install dependencies on the fly.

```bash
OPENAI_API_KEY='your-api-key' \
  uv run --with "pydantic-ai[examples]" \
  -m pydantic_ai_examples.pydantic_model
```

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

### Start the Realtime WebRTC server

Source: https://pydantic.dev/docs/ai/examples/realtime/realtime-webrtc/

Starts the uvicorn server for the example. Requires dependencies installed and OPENAI_API_KEY set. Open http://localhost:8000 and click Start call to test.

```bash
uv run --all-packages uvicorn pydantic_ai_examples.realtime_webrtc.app:app
```

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

### Run the example with pip

Source: https://pydantic.dev/docs/ai/examples/getting-started/pydantic-model/

Run the example using pip after dependencies are installed and environment variables are set.

```bash
python -m pydantic_ai_examples.pydantic_model
```

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

### Install examples extra dependencies with uv

Source: https://pydantic.dev/docs/ai/examples/setup/

Use this command to install the 'examples' optional dependency group when you have installed pydantic-ai via uv.

```bash
uv add "pydantic-ai[examples]"
```

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

### Run the realtime handoff example with uv

Source: https://pydantic.dev/docs/ai/examples/realtime/realtime-handoff/

Run the example using uv after installing dependencies and setting the OPENAI_API_KEY environment variable.

```bash
uv run -m pydantic_ai_examples.realtime_handoff
```

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

### Install uvicorn using pip

Source: https://pydantic.dev/docs/ai/integrations/ui/ag-ui/

Install uvicorn, an ASGI server, to run the examples.

```bash
pip install uvicorn
```

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

### Start a local Temporal server

Source: https://pydantic.dev/docs/ai/capabilities/durable_execution/temporal/

Install Temporal and start a development server locally, required for running the durable agent example.

```sh
brew install temporal
temporal server start-dev
```

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

### Run the realtime handoff example with pip

Source: https://pydantic.dev/docs/ai/examples/realtime/realtime-handoff/

Run the example using pip after installing dependencies and setting the OPENAI_API_KEY environment variable.

```bash
python -m pydantic_ai_examples.realtime_handoff
```

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

### Run the example with uv

Source: https://pydantic.dev/docs/ai/examples/realtime/realtime-voice/

Command to run the realtime voice example using uv. Requires dependencies installed and OPENAI_API_KEY set.

```bash
uv run -m pydantic_ai_examples.realtime_voice
```

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

### Run realtime text-to-audio example with uv

Source: https://pydantic.dev/docs/ai/examples/realtime/realtime-text-to-audio/

Run the example using uv. Requires OPENAI_API_KEY set and dependencies installed. Saves audio to realtime-response.wav.

```bash
uv run -m pydantic_ai_examples.realtime_text_to_audio "Tell me a fun fact about octopuses."
```

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

### Run stream_whales example with uv

Source: https://pydantic.dev/docs/ai/examples/streaming/stream-whales/

Run the example using uv after installing dependencies and setting environment variables.

```bash
uv run -m pydantic_ai_examples.stream_whales
```

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

### Run the example with pip

Source: https://pydantic.dev/docs/ai/examples/realtime/realtime-voice/

Command to run the realtime voice example using pip. Requires dependencies installed and OPENAI_API_KEY set.

```bash
python -m pydantic_ai_examples.realtime_voice
```

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

### Run the bank support example with pip

Source: https://pydantic.dev/docs/ai/examples/conversational-agents/bank-support/

Command to run the example using pip after dependencies are installed and environment variables are set.

```bash
python -m pydantic_ai_examples.bank_support
```

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

### Install Pylon harness with uv

Source: https://pydantic.dev/docs/ai/harness/pylon/

Install the Pylon harness and the OpenAI provider extra using uv. The OpenAI provider is required for the example; for other models, install that provider's extra instead.

```bash
uv add "pydantic-ai-harness[pylon]" "pydantic-ai-slim[openai]"
```

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

### start() - Initialize camera and start streaming

Source: https://pydantic.dev/docs/ai/examples/realtime/realtime-camera/

Requests camera and microphone access, sets up video element, connects, and starts mic, frames, and loop. Uses rear camera with ideal 1920x1080, downscaled to FRAME_W. Disables toggle button during start, sets running state, and updates UI.

```javascript
async function start() {
        if (starting) return;
        starting = true; toggleBtn.disabled = true;
        try {
          mediaStream = await navigator.mediaDevices.getUserMedia({
            audio: { channelCount: 1, echoCancellation: true, noiseSuppression: true },
            // Prefer the rear camera; frames are downscaled to FRAME_W before streaming to keep the
            // live session cheap.
            video: { facingMode: { ideal: 'environment' }, width: { ideal: 1920 }, height: { ideal: 1080 } },
          });
          video.srcObject = mediaStream;
          app.dataset.cam = 'on';
          await connect();
          ensurePlayback();
          startMic();
          startFrames();
          startLoop();
          running = true; toggleBtn.textContent = 'Stop'; toggleBtn.classList.add('stop'); watchBtn.disabled = false;
        } catch (err) {
          stop();
          throw err;
        } finally {
          starting = false; toggleBtn.disabled = false;
        }
      }
```

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

### Install examples extra dependencies with pip

Source: https://pydantic.dev/docs/ai/examples/setup/

Use this command to install the 'examples' optional dependency group when you have installed pydantic-ai via pip.

```bash
pip install "pydantic-ai[examples]"
```

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

### Run SQL Generation Example with uv

Source: https://pydantic.dev/docs/ai/examples/data-analytics/sql-gen/

Runs the example using uv. Use this command after setting up dependencies and environment variables.

```bash
uv run -m pydantic_ai_examples.sql_gen
```

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

### Run the camera agent with uv

Source: https://pydantic.dev/docs/ai/examples/realtime/realtime-camera/

Starts the local server using uv. Requires dependencies installed and environment variables set. Open http://localhost:8000 and select Start.

```bash
uv run -m pydantic_ai_examples.realtime_camera.app
```

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

### Configure LocalStack with manage_container=True

Source: https://pydantic.dev/docs/ai/harness/localstack/

Sets manage_container=True so a LocalStack Docker container is started for each run and stopped when the run ends. Requires Docker to be installed and running. The container's edge port (4566) is published on the host port from endpoint_url, and the capability waits for the health endpoint before starting. Each run gets its own container, so concurrent runs need distinct host ports or an externally managed instance.

```python
from pydantic_ai_harness import LocalStack

LocalStack(
    manage_container=True,
    image='localstack/localstack',
    container_env={'DEBUG': '1', 'PERSISTENCE': '1'},
    startup_timeout=120.0,
)
```

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

### Run the cancel_and_resume example with uv

Source: https://pydantic.dev/docs/ai/examples/streaming/cancellation/

Run the example using uv after dependencies are installed and environment variables are set. Press Esc to cancel a streaming turn, Ctrl+C or Ctrl+D to exit.

```bash
uv run -m pydantic_ai_examples.cancel_and_resume
```

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

### Run SQL Generation Example with pip

Source: https://pydantic.dev/docs/ai/examples/data-analytics/sql-gen/

Runs the example using pip-installed dependencies. Use this command after setting up dependencies and environment variables.

```bash
python -m pydantic_ai_examples.sql_gen
```

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

### Run realtime text-to-audio example with pip

Source: https://pydantic.dev/docs/ai/examples/realtime/realtime-text-to-audio/

Run the example using pip-installed dependencies. Requires OPENAI_API_KEY set and dependencies installed. Saves audio to realtime-response.wav.

```bash
python -m pydantic_ai_examples.realtime_text_to_audio "Tell me a fun fact about octopuses."
```

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

### Run stream_whales example with pip

Source: https://pydantic.dev/docs/ai/examples/streaming/stream-whales/

Run the example using pip after installing dependencies and setting environment variables.

```bash
python -m pydantic_ai_examples.stream_whales
```

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

### Install pydantic-ai-harness with pip

Source: https://pydantic.dev/docs/ai/harness/bubblewrap-sandbox/

Installs the package with the anthropic extra for the examples. Requires Linux with bwrap and user namespaces.

```bash
pip install "pydantic-ai-harness[anthropic]"
```

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

### Start Pydantic AI AG-UI Backend with uv

Source: https://pydantic.dev/docs/ai/examples/ag-ui/

Runs the AG-UI example backend using uv.

```bash
uv run -m pydantic_ai_examples.ag_ui
```

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

### Install DynamicWorkflow extra with uv

Source: https://pydantic.dev/docs/ai/harness/dynamic-workflow/

Install the pydantic-ai-harness package with the dynamic-workflow extra using uv. Alternative to pip for the same setup.

```bash
uv add "pydantic-ai-harness[dynamic-workflow]"
```

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

### Install PostHog harness with uv

Source: https://pydantic.dev/docs/ai/harness/posthog/

Install the PostHog harness and the OpenAI provider using uv. The OpenAI provider is required for the example; for other models, install that provider's extra instead.

```bash
uv add "pydantic-ai-harness[posthog]" "pydantic-ai-slim[openai]"
```

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

### Install PostHog harness with pip

Source: https://pydantic.dev/docs/ai/harness/posthog/

Install the PostHog harness and the OpenAI provider using pip. The OpenAI provider is required for the example; for other models, install that provider's extra instead.

```bash
pip install "pydantic-ai-harness[posthog]" "pydantic-ai-slim[openai]"
```

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

### Run SQL Generation Example with Custom Prompt (uv)

Source: https://pydantic.dev/docs/ai/examples/data-analytics/sql-gen/

Runs the example with a custom prompt using uv. The prompt is passed as a command-line argument.

```bash
uv run -m pydantic_ai_examples.sql_gen "find me errors"
```

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

### Run a Pydantic AI example with pip

Source: https://pydantic.dev/docs/ai/examples/setup/

Run any example module using python -m pydantic_ai_examples.<example_module_name> after installing pydantic-ai.

```bash
python -m pydantic_ai_examples.<example_module_name>
```
Tokens:
Raw