Models Settings in Pieces Desktop

Models Settings

Manage AI models and model preferences. Configure processing modes, set up local model runtime with Ollama, and control which AI models are available for use in Pieces.

To access Models settings, click your User Profile in the top left, then hover over Settings and select Models.

Models settings showing Model Capabilities, Local Model Runtime, and Model Management sections

Model Capabilities

Configure how Pieces processes your materials using machine learning resources. Choose between Cloud, Local, or Blended processing modes based on your performance and privacy preferences.

Processing Mode

Choose how Pieces processes your materials using machine learning resources. You can select Cloud, Local, or Blended processing modes based on your performance and privacy preferences.

Click your `User Profile` in the top left, then hover over `Settings` and select `Models`. In the *Model Capabilities* section, find the "Processing Mode" option showing your current mode (e.g., "Processing Mode: Blended"). Click the `Dropdown Arrow` next to the current processing mode to open the mode options. Choose from the following options: * **Cloud**: Optimized for speed, but all data gets sent to the cloud for processing * **Local**: Most processing happens locally before reaching out to the selected model, providing better privacy * **Blended**: Uses a combination of local and cloud resources, balancing speed and privacy Selecting the processing mode that best fits your security and performance needs ensures that Pieces processes your materials in the most efficient or privacy-conscious way possible.

Local Model Runtime

Set up and manage Ollama for local model processing. Ollama allows you to run AI models locally on your device without sending data to the cloud.

Ollama Status

Check if Ollama is installed, activated, and ready to use. Ollama must be installed and running for local model processing to work.

Click your `User Profile` in the top left, then hover over `Settings` and select `Models`. Scroll down to the *Local Model Runtime* section. View the Ollama status: * **Activated and Ready**: Shows "Ollama is Activated and Ready" with a green checkmark and version number (e.g., "Version: 0.5.5") * **Not Installed**: Shows options to install Ollama * **Not Running**: Shows options to start Ollama

Installing Ollama

If Ollama is not installed, you can install it directly from the Models settings page.

Click your `User Profile` in the top left, then hover over `Settings` and select `Models`. Scroll down to the *Local Model Runtime* section. If Ollama is not installed, click the `Install` button or link to download and install Ollama. Follow the installation prompts to complete the Ollama installation. Once installed, Ollama will automatically activate and be ready to use.

Model Management

Enable or disable models to control which AI models are available for use in Pieces. Disabled models will not appear in chat or other features. Models disabled by an organization will not appear here.

Understanding Model Management

Pieces supports a wide variety of AI models from different providers. By default, the most popular models are enabled, but you can customize which models are available based on your needs and preferences.

Viewing Enabled Models

See how many models are currently enabled and which providers they belong to.

Click your `User Profile` in the top left, then hover over `Settings` and select `Models`. Scroll down to the *Model Management* section. At the top of the *Model Management* section, you'll see a count showing how many models are enabled out of the total available (e.g., "3 of 68 models enabled").

Searching Models

Search for specific models or providers to quickly find what you're looking for.

Click your `User Profile` in the top left, then hover over `Settings` and select `Models`. In the *Model Management* section, find the search bar with the placeholder "Search models...". Type the name of a model or provider in the search bar to filter the list of available models.

Enabling or Disabling Models

Control which models are available by enabling or disabling them individually or by provider.

Click your `User Profile` in the top left, then hover over `Settings` and select `Models`. In the *Model Management* section, find the model provider you want to configure (e.g., OpenAI, Anthropic, Google, Microsoft, Meta, IBM). Click the arrow next to the provider name to expand and see all available models from that provider. Use the toggle switch on the right side of each model or provider to enable or disable it. Green indicates enabled, gray indicates disabled.

Models settings showing how to enable a model using the toggle switch

Deleting Local Models

Remove local models that you no longer need. This frees up storage space on your device.

Click your `User Profile` in the top left, then hover over `Settings` and select `Models`. In the *Model Management* section, find the local model you want to delete. Local models are typically from providers like Ollama or other local runtime environments. Click the delete icon or option next to the local model to remove it from your device.

Models settings showing how to delete a local model

Enabling All Models

Quickly enable all available models at once.

Click your `User Profile` in the top left, then hover over `Settings` and select `Models`. In the *Model Management* section, find the "Enable All" option with a checkmark icon. Click `Enable All` to enable all available models. This will make all models from all providers available for use.

Model Providers

Pieces supports models from multiple providers. Each provider shows how many of its models are enabled:

OpenAI: GPT models and other OpenAI models
Anthropic: Claude models
Google: Gemini and other Google models
Microsoft: Azure OpenAI and other Microsoft models
Meta: Llama and other Meta models
IBM: Watson and other IBM models

Each provider can be expanded to see individual models, and you can enable or disable models individually or toggle the entire provider on or off.

Compatible Models

For a complete list of all compatible models and their capabilities, see the Compatible LLMs documentation. This page provides detailed information about cloud models, local models, and which models work best for different use cases.

Free vs Pro Model Access

Model access depends on your plan. Free users get full access to local models and limited usage of cloud models. Pieces Pro subscribers get unlimited access to premium cloud LLMs.

Free Plan — Models

The free plan includes:

Local models (Ollama): Full access with no usage limits. Download and run models like Llama, Gemma, Phi, and others entirely on-device.
Cloud models: Limited usage. Access to select cloud models with usage caps—ideal for getting started and trying AI-assisted development.
Basic AI features: Snippet management, local storage, and community support.

Model availability and usage limits may vary. See the [Cloud Models](/products/large-language-models/cloud-models) page for the latest list and plan-specific availability.

Plan Comparison

Feature	Free	Pieces Pro
Local models (Ollama)	Full access, no limits	Full access, no limits
Cloud / premium models	Limited usage	Unlimited access
Long-term memory	Standard	Infinite
Deep Study reports	—	✓

For pricing and upgrade options, see Pieces Pro.

Next Steps

Now that you understand how to manage models, learn about Long-Term Memory to configure memory preferences, or explore Model Context Protocol (MCP) to integrate Pieces with other tools. For a complete list of available models, see Compatible LLMs.