Choose Your AI Model in Ask Lightspeed

Not every question needs the same model. A quick fact check doesn't require the same horsepower as a detailed strategy memo, and a fast draft shouldn't cost you the wait time of a deep-reasoning response. Starting today, you're in control.

A new model selector sits right in the Ask Lightspeed chat header, alongside the existing tone selector. One click switches between three Claude models, and your choice is remembered across sessions.

The models

• Sonnet 4.6 — Fast & capable — The default. Sonnet strikes the best balance between speed and quality for everyday tasks: drafting emails, answering questions, brainstorming ideas, and general-purpose work. Most users will want this most of the time.
• Opus 4.6 — Best for complex tasks — Anthropic's most powerful model. Choose Opus when you need deeper reasoning, nuanced analysis, long-form strategy documents, or when working through multi-step problems that require careful thought.
• Haiku 4.5 — Quick & lightweight — The fastest option. Haiku is ideal for simple lookups, quick calculations, short answers, and high-volume interactions where speed matters more than depth.

How it works

The model selector appears as a row of pills below the tone selector in the Ask Lightspeed chat header. Each pill shows the model name and a short description of what it's best for. Click to switch — the selection takes effect on your very next message.

Your model preference is saved automatically and persists across sessions, so you don't need to re-select each time you open Ask Lightspeed.

Good to know

• Model selection applies to Ask Lightspeed only — other tools like the Response Assistant and Draft Assistant continue to use the organization-configured default
• All three models have full access to your organization's knowledge base, rated examples, and Shopify data (if connected)
• Switching models mid-conversation is fully supported — your conversation history carries over seamlessly