Hi Teable team and everyone here!
I really like the idea of having Lite, Smart, and Ultra presets. They make the App Builder easier to use without forcing everyone to understand model settings.
But I feel the current OpenAI-only selection is a little too restrictive. AI models have different strengths, and sometimes switching models can save more time than rewriting the same prompt five different ways.
For example, if a model keeps misunderstanding a requirement or struggling with a code change, I’d love to try Claude Haiku or Sonnet instead of repeatedly trying to make the same model understand. Different models approach problems differently, and that flexibility can be genuinely useful when building real applications.
Cost is another reason. For standard API pricing, GPT-6 Luna costs $0.10 per million input tokens and $0.50 per million output tokens. Claude Haiku 5.5 has the same rates for prompts up to 100K tokens. At the higher end, GPT-6.1 Sol and Claude Sonnet 5.5 both cost $2 per million input tokens and $10 per million output tokens. Yet they can behave differently on the same task. OpenAI pricing · Anthropic pricing.
I’m not suggesting that one model is universally better than another. I just want the freedom to choose the one that works best for a particular task. It could help us debug stubborn issues, get a second opinion, reduce unnecessary iterations, and balance quality with cost.
I also remember when the App Builder offered a selection of 180+ models. I’d genuinely like to understand why that option was removed. Was it due to reliability, maintenance, cost, or another product decision? I understand that supporting every model isn’t practical, and I’m not asking for the entire catalog back.
Could we instead have a small, curated selection of frontier models from multiple providers within the existing presets? Teable could test them for coding quality, tool compatibility, reliability, and cost before adding them.
What I’m requesting
-
Expand Lite, Smart, and Ultra with a few carefully selected models from other providers, including Anthropic.
-
Allow model switching between approved options when a task isn’t going well.
-
Keep it simple with a curated list, rather than restoring all 180+ models.
-
Explain why the previous model selection was removed and whether broader model support is being considered.
I’d love to hear what other builders think. Does anyone else want this flexibility? If you’ve had a model struggle with a task that another model handled better, please share your experience.
I believe this could make Teable’s App Builder more useful without making it more complicated. I’d really appreciate the community’s support in getting this considered.