AI Models

ClawHQ gives you access to a curated selection of bundled AI models. Choose the model that best fits your use case, compare capabilities side by side, and switch models with a single click.

Current Model Display

The top of the AI Models page shows your currently active model in a prominent card. The card displays:

• Model name — The human-readable name of the active model.
• Context window — The maximum number of tokens the model can process in a single conversation turn (e.g., 8K, 32K, 128K tokens).
• Performance badge — A visual indicator showing whether the model is categorized as Fast, Balanced, or Powerful.

The current model is the one that all deployed agents use to generate responses. Switching the model changes the underlying AI for every agent on your instance simultaneously.

Available Models

Below the current model, the page displays a grid of all available models as individual cards. Each model card includes:

• The model name and a brief description of its strengths.
• The context window size.
• A performance badge (Fast, Balanced, or Powerful).
• A "Switch to this model" button (or "Active" if it is already the current model).

Models are sorted by capability tier, with the most powerful models listed first. Each card is color-coded by its performance category to make scanning the list faster.

Model Comparison

The model comparison feature lets you evaluate two or more models side by side before making a switch. Select the models you want to compare and the comparison view displays them in adjacent columns with the following attributes:

• Capabilities — Checkmarks indicating support for chat, analysis, code generation, vision, creative writing, and fast inference.
• Strengths — A short text summary of what the model excels at.
• Speed rating — A relative indicator (1-5 scale) of how quickly the model generates responses.
• Context window — The token limit for the model.

The comparison table highlights differences between the selected models, making it straightforward to identify which model offers the best fit for your workload.

Model Switches

Starter plan users can perform up to 5 model switches per calendar month. The current count of switches used is displayed on the page alongside a progress indicator. Pro and Ultra plan users have unlimited model switches.

Each switch counts as one use regardless of whether you switch to a new model or revert to a previous one. The counter resets on the first day of each billing cycle.

Switch Confirmation

When you click "Switch to this model," a confirmation dialog appears before the change is applied. The dialog provides a direct comparison between your current model and the model you are switching to, showing:

• Both model names side by side.
• Context window comparison (e.g., "8K → 128K").
• Performance badge comparison (e.g., "Fast → Powerful").
• A note about the expected downtime during the switch.
• For Starter users, the remaining switch count after this action.

You must explicitly confirm the switch by clicking the "Confirm Switch" button. The process typically completes within 30 seconds, during which a progress indicator is displayed.

Model Recommendation Engine

Not sure which model to choose? The recommendation engine analyzes your use case and suggests the best model for your needs. Click the "Get Recommendation" button and answer a few quick questions:

• What is the primary purpose of your agents? (Customer support, research, content creation, code assistance, general purpose)
• How important is response speed vs. response quality?
• Do your conversations tend to be short exchanges or long, multi-turn discussions?
• Do you need vision or image analysis capabilities?

Based on your answers, the engine highlights the recommended model with a brief explanation of why it is the best match. You can then switch to the recommended model directly from the results view.

Model Detail Dialog

Clicking on any model card opens a detail dialog with the full specification sheet for that model. The dialog includes:

• Complete model name and version identifier.
• Detailed capability list with descriptions of each supported feature.
• Context window size with an explanation of what it means in practice.
• Speed benchmarks relative to other available models.
• Ideal use cases and scenarios where the model excels.
• Known limitations or trade-offs compared to other models.

Performance Badges

Every model in ClawHQ is assigned one of three performance badges to help you quickly assess its characteristics:

• Fast — Optimized for low-latency responses. Ideal for real-time customer support and high-volume chat applications where speed is the priority.
• Balanced — A middle ground between speed and capability. Suitable for most general-purpose workloads including customer service, content generation, and light analysis.
• Powerful — The highest capability tier. Best for complex reasoning, code generation, long-document analysis, and tasks that require the deepest understanding.

Related Documentation

• Dashboard Overview — See your model status at a glance.
• Agents — Understand how agents use the active model.
• Chat — Test model performance in real-time conversations.
• VPS Management — Monitor VPS impact during model switches.