The right model for every task
Every model is included with Modurian — pre-configured and ready on sign-in. Choose Auto or pick manually on Pro.
No API keys required. Modurian hosts model access for you. Free accounts use Auto on fast-tier models; Pro unlocks all models, manual selection, and reasoning depth controls. See Included models in Resources for details.
| Model | Best for |
|---|---|
| Auto (Smart Route)Recommended | Let Modurian pick the best model for each message |
| DeepSeek V4 ProPremium | Architecture, debugging, complex logic, planning |
| DeepSeek V4 FlashFast | Fast coding, simple tasks, routine edits |
| MiniMax M3Premium | Latest MiniMax — coding, reasoning, and vision |
| MiniMax M2.7Premium | Code generation and multi-step builds |
| M2.7 HighspeedFast | Quick refactors and simple coding |
Auto routing examples
- “Architect a blog system” → premium tier (architecture)
- “Fix the broken login” → premium tier (debugging)
- “Add a blue button” → fast tier (simple edit)
How routing works
Pick Auto — or take the wheel on Pro
Model choice is a product decision, not an ops chore. Here is how Modurian thinks about it.
Auto routing vs manual pick
Auto watches the intent of each message — architecture questions and hard bugs tend to escalate to stronger reasoning models, while quick copy edits stay on fast flash tiers. On Pro you can override Auto and pin a model for the whole session when you know you need maximum reasoning depth or maximum speed.
Context windows that fit real projects
Long-context models keep multi-file projects coherent without constant re-pasting. Modurian also maintains project memory and rules so agents remember your stack preferences across turns. That matters more than raw model marketing numbers when you are iterating for hours.
Vision and Snap-to-Build
Pro unlocks vision-capable models for screenshot understanding. Snap-to-Build uses that stack to rebuild UI from an image. Attach reference screenshots or PDFs when a verbal description alone would underspecify the layout.
FAQ
Frequently asked questions
Which model should I use as a beginner?
Leave Auto on. Free accounts already route through capable flash models. Upgrade to Pro when you want the full catalog or need vision for Snap-to-Build.
Does switching models mid-session break context?
No. Conversation history and the virtual filesystem stay intact. Only the model that answers the next message changes.
How do credits relate to models?
Heavier models and longer contexts consume more credits per turn. Auto tries to balance quality and cost; pinning a frontier model intentionally costs more.