### Answer The issue you're experiencing with incorrect date calculations is a known challenge when relying on the LLM to perform date arithmetic. While

{{current_time_Europe/London}}

correctly provides the current timestamp (e.g., "Thursday, March 28, 2024 at 11:46:04 PM GMT") , the LLM itself may still make errors when calculating relative dates like "tomorrow." [(1) ]() Here are some approaches to address this: 1. Use the Calendar Variable Instead For date-related tasks, consider using

{{current_calendar_Europe/London}}

which provides a 14-day calendar with dates already calculated. This gives the LLM explicit date references rather than requiring it to compute them. 2. Lower Temperature Doesn't Guarantee Accuracy While you've set temperature to 0.00 (which makes responses more deterministic) , this doesn't prevent the LLM from making calculation errors—it just makes it more likely to repeat the same error consistently. [(2) ]() 3. Pre-Calculate Dates Externally For critical date accuracy, consider passing pre-calculated dates as dynamic variables via your webhook or API call. For example, pass

{{tomorrow_date}}

already formatted as "04/03/2026" rather than asking the LLM to calculate it. [(1) ]() When using the Create Phone Call API, you can set these in `retell_llm_dynamic_variables`:

json
{
  "today_date": "03/03/2026",
  "tomorrow_date": "04/03/2026"
}

[(2) ]() 4. Strengthen Prompt Instructions In your prompt, explicitly state the current date and provide examples:

Today is {{current_time_Europe/London}}.
When I say "tomorrow", calculate exactly one day forward.
Example: If today is 03/03/2026, tomorrow is 04/03/2026.

The fundamental limitation is that LLMs can struggle with arithmetic and date calculations even with explicit instructions. Pre-calculating dates and passing them as dynamic variables is the most reliable solution. [(1) ]()

​ To learn more, see the sources I considered: [[1] Dynamic Variables](https://docs.retellai.com/build/dynamic-variables) [[2] Create Retell LLM](https://docs.retellai.com/api-references/create-retell-llm) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer Testing a different temperature is unlikely to solve your date calculation issue. Here's why: Temperature affects randomness, not accuracy. [(1) ]() - Temperature 0.0–0.3 produces highly consistent, deterministic responses—ideal for data collection and function calling - However, this just means the LLM will consistently repeat the same error rather than vary its mistakes For tool calling and data-heavy tasks, a lower temperature is actually recommended. [(2) ]() The core problem: LLMs inherently struggle with arithmetic and date calculations regardless of temperature. Raising the temperature would introduce more variability but wouldn't improve calculation accuracy. Better solutions: 1. Pre-calculate dates externally and pass them as dynamic variables (e.g.,

{{tomorrow_date}}

= "04/03/2026") via your webhook or API 2. Use

{{current_calendar_Europe/London}}

which provides a 14-day calendar with dates already calculated, removing the need for the LLM to compute them Pre-calculating dates and injecting them as dynamic variables remains the most reliable approach for date-critical workflows.

​ To learn more, see the sources I considered: [[1] Configure LLM Options](https://docs.retellai.com/build/llm-options) [[2] Create Retell LLM](https://docs.retellai.com/api-references/create-retell-llm) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.