LLMTool
Listing of Parameter Types »Details
- LLMTool is typically used to provide exact answers or up-to-date information to a large language model (LLM).
- Using LLMTool modifies an LLM prompt to explain how to use the tool, what it does and its arguments.
- The LLM can use the tool by specifying its "name", which should be descriptive of the actual tool function for triggering LLM usage effectively. Extra details about the tool can be given in "description".
- Tool calls are parsed into an LLMToolRequest and the evaluation result is sent to the LLM as a string.
- The format of params and fun is analogous to APIFunction .
- Valid forms of params include:
-
"name" string parameter"name"type typed parameter{spec1,spec2,…} list of parameters
- Valid forms of type include:
-
"form" any form supported by Interpreterassoc an association giving details of how to treat the parametertfun any function to be applied to the string
- Possible elements in assoc include:
-
"Interpreter" "String" interpreter type"Required" True whether the parameter is required
Examples
open all close allBasic Examples (2)
Define a tool for counting the number of characters in a string:
tool = LLMTool["countcharacter", "string", StringLength[#string]&]Use the tool in a call to an LLM:
LLMSynthesize["How many letters in the word \"characters\" (use the tool) ?", LLMEvaluator -> <|"Tools" -> tool|>]Define a tool for querying the population of cities by their name and describe it for the LLM:
tool = LLMTool[{"cityPopulationFinder", "Query WolframAlpha to get the population of a city from its name."}, "city" -> "City", #city["Population"]&]Use the tool in a call to an LLM:
LLMSynthesize["Use the tool to write a few sentences about Chicago.", LLMEvaluator -> <|"Tools" -> tool|>]Scope (3)
Define a tool for computing power series expansion using multiple different inputs interpretations:
series = LLMTool[{"series_expand", "compute taylor series of a Wolfram Expression"},
{"input" -> "Expression", "order" -> "Integer", "variable" -> "Expression"}, ToString[Normal@Series[#input, {#variable, 0, #order}], InputForm]&]Ask the LLM a question requiring it to query a series expansion for the answer:
LLMSynthesize["What is the tenth order coefficient of w*Sin[w] taylor expansion?", "CompletionText", LLMEvaluator -> LLMConfiguration[<|"Tools" -> series|>]]Define a tool to expose the TravelDistance function, interpreting parameter as a list of cities:
trip = LLMTool[{"CityTravelDistance", "Gives the estimated distance traveling to each city in the list"}, {"cityList" -> DelimitedSequence["City", ","]}, TravelDistance[#cityList]&]Find shortest travel path:
LLMSynthesize["Going from NewYork to Boston is it shorter to pass by Springfield or Providence ?", LLMEvaluator -> <|"Tools" -> trip|>]Define a tool with some prompting for running Wolfram Language code:
tool = LLMTool[{"runCode", "Run Wolfram Language code"}, "code", If[ChoiceDialog["Run this code?
" <> #code], ToString[ToExpression[#code], InputForm], "Code not allowed"]&]Run an LLM with the tool, returning all the evaluation properties:
data = LLMSynthesize["Which is closer to Boston: San Francisco or London? Use the tool as you write your response.", All, LLMEvaluator -> <|"Tools" -> tool|>];Get the completion written by the LLM with prompting and tool calls removed:
data["CompletionText"]Get the completion with tool calls preserved:
data["CompletionToolsText"]Get the LLMToolRequest and LLMToolResponse objects lists:
data["ToolRequests"]data["ToolResponses"]History
Introduced in 2023 (13.3)
Text
Wolfram Research (2023), LLMTool, Wolfram Language function, https://reference.wolfram.com/language/ref/LLMTool.html.
CMS
Wolfram Language. 2023. "LLMTool." Wolfram Language & System Documentation Center. Wolfram Research. https://reference.wolfram.com/language/ref/LLMTool.html.
APA
Wolfram Language. (2023). LLMTool. Wolfram Language & System Documentation Center. Retrieved from https://reference.wolfram.com/language/ref/LLMTool.html
BibTeX
@misc{reference.wolfram_2026_llmtool, author="Wolfram Research", title="{LLMTool}", year="2023", howpublished="\url{https://reference.wolfram.com/language/ref/LLMTool.html}", note=[Accessed: 13-August-2026]}
BibLaTeX
@online{reference.wolfram_2026_llmtool, organization={Wolfram Research}, title={LLMTool}, year={2023}, url={https://reference.wolfram.com/language/ref/LLMTool.html}, note=[Accessed: 13-August-2026]}