Models
Models are the AI Alfred can use. Your model choice affects answer quality, speed, usage cost, supported inputs, and how much context Alfred can consider.
How to choose
Section titled “How to choose”Use a stronger model for hard reasoning, coding, detailed writing, complex file analysis, and high-stakes planning.
Use a faster or lower-cost model for quick questions, drafts, simple summaries, and high-volume work.
Availability
Section titled “Availability”Model availability depends on your plan.
| Model | Best for | Inputs | Availability |
|---|---|---|---|
| GPT-5 | Balanced advanced work, reasoning, images, professional writing | Text, images | Elite and above |
| GPT-5-MINI | Faster OpenAI responses and lighter work | Text, images | Power and above |
| Claude-Sonnet | Strong coding, analysis, and careful reasoning | Text, images | Elite and above |
| Claude-Haiku | Fast everyday work with solid reasoning | Text, images | Power and above |
| Gemini-Pro | Large context, complex analysis, multimodal work | Text, images, audio, video | Elite and above |
| Gemini-Flash | Fast, cost-effective general work | Text, images, audio, video | All plans |
| Gemini-Flash-Lite | Lowest-cost conservation mode | Text, images, audio, video | All plans |
| DeepSeek-Flash | Fast text-only work | Text | All plans |
| DeepSeek-Pro | Higher-quality text-only work | Text | All plans |
Plan names and exact availability can change over time. The Dashboard model picker is the source of truth for your account because it shows what your current plan can use.
Automatic model switching
Section titled “Automatic model switching”Your selected model is a preference. When your usage gets high, Alfred may temporarily switch to a more cost-effective model to keep your daily allowance available longer.
You may see Alfred mention that he has switched models. This usually means you are approaching a usage threshold, not that your preferred model setting was removed.
Input support
Section titled “Input support”Some models can work with images, audio, video, or large context better than others. AI models that cannot process images are given a Vision Agent that will mediate vision tasks on its behalf.
Cost and speed
Section titled “Cost and speed”More capable models often use more allowance. Larger prompts, long files, web research, generated outputs, and tool use can also increase usage. For simple work, a cheaper model can be the better experience because it is faster and lets you do more before reaching your limit.
See Usage for how allowance works.