Model Selection in Rolai
Rolai gives you the flexibility to choose from a wide range of leading AI models, including those from OpenAI, Claude, Gemini, and our own Rolai model, so you can match the right model to your specific task. Whether you need advanced reasoning, creative content generation, fast summaries, or cost-effective performance, you can select the model that best fits your workflow directly within the platform. This flexibility ensures you always have the optimal balance of speed, accuracy, and capability for every use case.
Model Selection in Rolai
Rolai lets you choose from a curated set of leading AI models so you can match the right model to each task directly from the Default Models tab in the model picker. You can switch models at any time, or use the My Agents tab to run conversations through your own preconfigured agents.
When you open the model picker you’ll see:
Default Models – all ready‑to‑use foundation models available in your workspace.
My Agents – your saved, custom Rolai agents built on top of these models, plus a Create New Agent button in the top‑right to add new ones.
Under Default Models, you’ll see groups of models from different providers:
OpenAI (GPT) – GPT‑5.2 Instant/Thinking/Pro, GPT‑5.1 Instant/Thinking, GPT‑5 Instant/Thinking/Pro, GPT‑4.1, and GPT‑4o. Use the “Thinking” and “Pro” variants for deeper reasoning and complex tasks, and the “Instant” options when speed and cost are more important.
Anthropic (Claude) – Claude 4.6 Opus and Sonnet, Claude 4.6 Sonnet Thinking, Claude 4 Sonnet, and the Claude 4.5 family (Sonnet, Sonnet Thinking, Haiku, Haiku Thinking). Choose Opus or Sonnet Thinking for advanced reasoning and Haiku variants for quick, lightweight work.
Google (Gemini) – Gemini 3.1 Pro, Gemini 3 Pro, Gemini 2.5 Pro, Gemini 2.5 Flash, and Gemini 3 Flash. Pro models focus on depth and analysis, while Flash models prioritize fast responses at lower cost.
As a simple rule of thumb, pick a Thinking/Pro/Opus model when you care most about reasoning quality, and an Instant/Haiku/Flash model when you care most about speed and efficiency. If you find yourself reusing the same setup repeatedly, save it as an agent under My Agents so the model choice, instructions, and tools are all preserved for next time.
Rolai’s Own Model
Rolai: Optimized for simple, low-latency tasks like document retrieval or snippet extraction, ensuring speed and reliability.
Additional Support
If you need any assistance, email support@rolai.com or use the Intercom chat feature in the app for direct help from our support team.