diff --git a/docs/docs/audio/speech.md b/docs/docs/audio/speech.md
index 2f7f4bb68..dbbbc1c48 100644
--- a/docs/docs/audio/speech.md
+++ b/docs/docs/audio/speech.md
@@ -95,6 +95,12 @@ ai_say('hello')
       Your browser does not support the audio element.
     </audio>
 
+## Choosing a voice
+
+Both `speak` and `@speech` accept a `voice` parameter that allows you to choose from a variety of voices. You can preview the available voices [here](https://platform.openai.com/docs/guides/text-to-speech/voice-options).
+
+```python
+
 ## Saving audio files
 
 The result of the `speak` function and `@speech` decorator is an audio stream. You can save this stream to disk like this:
@@ -102,4 +108,8 @@ The result of the `speak` function and `@speech` decorator is an audio stream. Y
 ```python
 audio = marvin.speak("Hello, world!")
 audio.stream_to_file("hello_world.mp3")
-```
\ No newline at end of file
+```
+
+
+## Model parameters
+You can pass parameters to the underlying API via the `model_kwargs` arguments of `speak` and `@speech`. These parameters are passed directly to the respective APIs, so you can use any supported parameter.
\ No newline at end of file
diff --git a/docs/docs/text/classification.md b/docs/docs/text/classification.md
index 0f496eda3..812b5b913 100644
--- a/docs/docs/text/classification.md
+++ b/docs/docs/text/classification.md
@@ -166,6 +166,9 @@ assert issue == IssueType.BUG
 
 While convenient for certain scenarios, it's recommended to use the `classify` function for its greater flexibility and broader application range.
 
+## Model parameters
+You can pass parameters to the underlying API via the `model_kwargs` argument of `classify` or `@classifier`. These parameters are passed directly to the API, so you can use any supported parameter.
+
 ## Best practices
 
 1. **Choosing the right labels**: Opt for labels that are mutually exclusive and collectively exhaustive for your classification context. This ensures clarity and prevents overlaps in categorization.
diff --git a/docs/docs/text/extraction.md b/docs/docs/text/extraction.md
index 10d83a3a8..5d0fbbd26 100644
--- a/docs/docs/text/extraction.md
+++ b/docs/docs/text/extraction.md
@@ -152,3 +152,6 @@ marvin.cast('Mass.', to=str, instruction="The state's abbreviation")
 # MA
 ```
 
+
+## Model parameters
+You can pass parameters to the underlying API via the `model_kwargs` argument of `extract`. These parameters are passed directly to the API, so you can use any supported parameter.
diff --git a/docs/docs/text/functions.md b/docs/docs/text/functions.md
index 786f1cf5d..b57ad8c2c 100644
--- a/docs/docs/text/functions.md
+++ b/docs/docs/text/functions.md
@@ -290,4 +290,8 @@ summarize_url('https://www.askmarvin.ai')
 ```
 
 ## Running a function
-Running a function is quite simple: just call it like you would any other function! The LLM will generate the output based on the function's definition and the provided inputs. Remember that no source code is generated or executed, so every call to the function will be handled by the LLM. You can use caching or other techniques to improve performance if necessary.
\ No newline at end of file
+Running a function is quite simple: just call it like you would any other function! The LLM will generate the output based on the function's definition and the provided inputs. Remember that no source code is generated or executed, so every call to the function will be handled by the LLM. You can use caching or other techniques to improve performance if necessary.
+
+
+## Model parameters
+You can pass parameters to the underlying API via the `model_kwargs` argument of `@fn`. These parameters are passed directly to the API, so you can use any supported parameter.
diff --git a/docs/docs/text/generation.md b/docs/docs/text/generation.md
index 16facdbb5..e5d092c12 100644
--- a/docs/docs/text/generation.md
+++ b/docs/docs/text/generation.md
@@ -124,6 +124,10 @@ Data generation relies even more on instructions than other Marvin tools, as the
 
 Instructions are freeform natural language and can be as general or specific as you like. The LLM will do its best to comply with any instructions you give.
 
+
+## Model parameters
+You can pass parameters to the underlying API via the `model_kwargs` argument of `generate`. These parameters are passed directly to the API, so you can use any supported parameter.
+
 ## Caching
 
 Normally, each `generate` call would be independent. For some prompts, this would mean that each call produced very similar results to other calls. That would mean that generating, say, 10 items in a single call would produce a much more varied and high-quality result than generating 10 items in 5 calls of 2 items each.
@@ -207,4 +211,4 @@ The first and second tabs both show high-quality, varied results. The third tab
             'Los Angeles',
             'Houston'
         ]
-        ```
\ No newline at end of file
+        ```
diff --git a/docs/docs/text/transformation.md b/docs/docs/text/transformation.md
index fa57f1088..b40c57bd4 100644
--- a/docs/docs/text/transformation.md
+++ b/docs/docs/text/transformation.md
@@ -111,4 +111,7 @@ class Location(marvin.Model):
 
 Location('CHI')
 # Location(city="Chicago", state="IL")
-```
\ No newline at end of file
+```
+
+## Model parameters
+You can pass parameters to the underlying API via the `model_kwargs` argument of `cast` or `@model`. These parameters are passed directly to the API, so you can use any supported parameter.
diff --git a/docs/docs/vision/captioning.md b/docs/docs/vision/captioning.md
index cf9abe54e..855d21146 100644
--- a/docs/docs/vision/captioning.md
+++ b/docs/docs/vision/captioning.md
@@ -40,3 +40,8 @@ Marvin can use OpenAI's vision API to process images as inputs.
   </p>
 </div>
 
+
+
+
+## Model parameters
+You can pass parameters to the underlying API via the `model_kwargs` argument of `caption`. These parameters are passed directly to the API, so you can use any supported parameter.
diff --git a/docs/docs/vision/classification.md b/docs/docs/vision/classification.md
index 3179c39cc..ce3c81708 100644
--- a/docs/docs/vision/classification.md
+++ b/docs/docs/vision/classification.md
@@ -55,3 +55,9 @@ The `marvin.beta.classify` function is an enhanced version of `marvin.classify`
         assert animal == 'dog'
         assert dry_or_wet == 'wet'
         ```
+
+
+
+
+## Model parameters
+You can pass parameters to the underlying API via the `model_kwargs` and `vision_model_kwargs` arguments of `classify`. These parameters are passed directly to the respective APIs, so you can use any supported parameter.
\ No newline at end of file
diff --git a/docs/docs/vision/extraction.md b/docs/docs/vision/extraction.md
index 56f8fd9d7..97b1df30a 100644
--- a/docs/docs/vision/extraction.md
+++ b/docs/docs/vision/extraction.md
@@ -48,3 +48,6 @@ The `marvin.beta.extract` function is an enhanced version of `marvin.extract` th
         ```python
         result == ["Pembroke Welsh Corgi", "Yorkshire Terrier"]
         ```    
+
+## Model parameters
+You can pass parameters to the underlying API via the `model_kwargs` and `vision_model_kwargs` arguments of `extract`. These parameters are passed directly to the respective APIs, so you can use any supported parameter.
diff --git a/docs/docs/vision/transformation.md b/docs/docs/vision/transformation.md
index 8ad0427b3..0e2c1e19b 100644
--- a/docs/docs/vision/transformation.md
+++ b/docs/docs/vision/transformation.md
@@ -113,3 +113,6 @@ If the target type isn't self-documenting, or you want to provide additional gui
         ```python
         assert missing_items == ["eggs", "oranges"]
         ```
+
+## Model parameters
+You can pass parameters to the underlying API via the `model_kwargs` and `vision_model_kwargs` arguments of `cast`. These parameters are passed directly to the respective APIs, so you can use any supported parameter.
\ No newline at end of file
diff --git a/src/marvin/ai/audio.py b/src/marvin/ai/audio.py
index 53bfa9576..7cead0668 100644
--- a/src/marvin/ai/audio.py
+++ b/src/marvin/ai/audio.py
@@ -80,7 +80,12 @@ def speak(
     return response
 
 
-def speech(fn: Optional[Callable] = None, *, voice: Optional[str] = None) -> Callable:
+def speech(
+    fn: Optional[Callable] = None,
+    *,
+    voice: Optional[str] = None,
+    model_kwargs: Optional[dict] = None,
+) -> Callable:
     """
     Function decorator that generates audio from the wrapped function's return
     value. The voice used for the audio can be specified.
@@ -93,11 +98,11 @@ def speech(fn: Optional[Callable] = None, *, voice: Optional[str] = None) -> Cal
         Callable: The wrapped function.
     """
     if fn is None:
-        return partial(speech, voice=voice)
+        return partial(speech, voice=voice, model_kwargs=model_kwargs)
 
     @wraps(fn)
     def wrapper(*args, **kwargs):
         model = PythonFunction.from_function_call(fn, *args, **kwargs)
-        return speak(text=model.return_value, voice=voice)
+        return speak(text=model.return_value, voice=voice, model_kwargs=model_kwargs)
 
     return wrapper