Removal of Model Selection in Base44 Is Blocking Real Workflows

Hello Base44 team and community,

I’d like to raise a serious concern regarding a recent change on the Base44 platform.

Until recently, when building systems, we were able to choose the agent model (Sonnet, Opus, Gemini, ChatGPT). This worked extremely well for us — different models were used for different tasks, and our workflows were stable, predictable, and efficient.

Now, the platform has been changed so that model selection is no longer possible and everything runs in automatic mode only.

Unfortunately, this change is severely impacting our ability to work and has completely disrupted our plans and system architecture.

We understand that maintaining access to multiple model providers can be expensive, and that cost optimization is important. However, we are also paying a significant amount for this platform specifically to build professional systems. Removing such a critical control without an alternative is not a fair trade-off.

There were clearly better solutions available, for example:

• Different pricing tiers for different models

• Pay-per-use or credits per model

• Allowing advanced users to manually select models

• Keeping “auto mode” as an option, not the only choice

Different models excel at different tasks. Auto-selection cannot replace intentional architectural decisions, especially for complex systems.

Right now, this feels like a step backwards for power users and teams building serious products on Base44.

We strongly ask the team to reconsider this decision and bring back manual model selection, even if it requires higher-tier plans or additional costs.

Base44 is a powerful platform — we want to continue building on it, but we need the tools and control to do our work properly.

Thank you for listening.