The Model tab has four settings.
Provider
Which of your Tenant Settings → AI providers to use. Leave blank to auto-detect: the workspace's default provider if one is set, otherwise the first enabled provider with a valid key.
Model
Leave blank to use that provider's configured default model — the recommended choice for production, since it lets you upgrade the model for every logic block at once by changing the provider setting, rather than hunting down every step that hardcoded a model name.
Max output tokens
Defaults to 2048. Responses cut off past this limit come back with usage.truncated: true on the step output — check that field if output looks incomplete before assuming something else is wrong. Raising it helps for long generations, but a synchronous HTTP call to a logic block is hard-capped around 29 seconds by the API Gateway integration limit regardless of this setting; for long-running generations, trigger the block from an automation (up to 5 minutes) or invoke it over the WebSocket connection instead.
Temperature
Leave blank to use the recommended default, which is model-aware:
- Reasoning models default to 1 — the only value they accept, and the engine detects them by model name: OpenAI's o-series (
o1,o3,o4-mini, etc.), Kimi'skimi-k3, DeepSeek'sdeepseek-reasoner, and any model name containing-thinking. - Every other model defaults to 0.3 — deterministic-leaning, appropriate for most classify/extract/generate tasks.
A reasoning model with a manually-set temperature will fail
If you explicitly set a temperature (any value other than 1) and later switch the model to a reasoning model, the explicit value still wins over the auto-default — and the provider will reject the request with a 400 ("invalid temperature: only 1 is allowed for this model"). Clear the field back to blank when switching to a reasoning model, or set it to 1 explicitly.
There's no streaming toggle here — streaming is determined entirely by how the logic block is invoked (the WebSocket customapi_run action streams tokens live; the synchronous HTTP route does not), not by a step config field.