Skip to content

pydantic_ai.settings

ModelSettings

Bases: TypedDict

Settings to configure an LLM.

Includes only settings which apply to multiple models / model providers, though not all of these settings are supported by all models.

All types must be serializable using Pydantic.

Source code in pydantic_ai_slim/pydantic_ai/settings.py 
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
class ModelSettings(TypedDict, total=False):
    """Settings to configure an LLM.

    Includes only settings which apply to multiple models / model providers,
    though not all of these settings are supported by all models.

    All types must be serializable using Pydantic.
    """

    max_tokens: int
    """The maximum number of tokens to generate before stopping.

    Supported by:

    * Gemini
    * Anthropic
    * OpenAI
    * Groq
    * Cohere
    * Mistral
    * Bedrock
    * MCP Sampling
    * xAI
    """

    temperature: float
    """Amount of randomness injected into the response.

    Use `temperature` closer to `0.0` for analytical / multiple choice, and closer to a model's
    maximum `temperature` for creative and generative tasks.

    Note that even with `temperature` of `0.0`, the results will not be fully deterministic.

    Supported by:

    * Gemini
    * Anthropic
    * OpenAI
    * Groq
    * Cohere
    * Mistral
    * Bedrock
    * xAI
    """

    top_p: float
    """An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass.

    So 0.1 means only the tokens comprising the top 10% probability mass are considered.

    You should either alter `temperature` or `top_p`, but not both.

    Supported by:

    * Gemini
    * Anthropic
    * OpenAI
    * Groq
    * Cohere
    * Mistral
    * Bedrock
    * xAI
    """

    timeout: float | Timeout
    """Override the client-level default timeout for a request, in seconds.

    Supported by:

    * Gemini
    * Anthropic
    * OpenAI
    * Groq
    * Mistral
    * xAI
    """

    parallel_tool_calls: bool
    """Whether to allow parallel tool calls.

    Supported by:

    * OpenAI (some models, not o1)
    * Groq
    * Anthropic
    * xAI
    """

    tool_choice: ToolChoice
    """Control which function tools the model can use.

    See the [Tool Choice guide](../tools-advanced.md#tool-choice) for detailed documentation
    and examples.

    * `None` (default): Defaults to `'auto'` behavior
    * `'auto'`: All tools available, model decides whether to use them
    * `'none'`: Disables function tools; model responds with text only (output tools remain for structured output)
    * `'required'`: Forces tool use; excludes output tools so the agent cannot produce a final response when set statically
    * `list[str]`: Only specified tools; excludes output tools so the agent cannot produce a final response when set statically
    * [`ToolOrOutput`][pydantic_ai.settings.ToolOrOutput]: Specified function tools plus output tools/text/image

    Note: setting `'required'` or `list[str]` *statically* (via the `model_settings` argument
    of [`Agent.run`][pydantic_ai.Agent.run] or the agent's own `model_settings`) raises a
    `UserError`, because it would force a tool call on every step and prevent the agent from
    producing a final response. To vary `tool_choice` per step (e.g. force a tool on the
    first step only), return a callable from a capability's
    [`get_model_settings`][pydantic_ai.capabilities.AbstractCapability.get_model_settings] —
    those values are trusted to adapt across steps. For single API calls without an agent
    loop, use [`pydantic_ai.direct.model_request`][pydantic_ai.direct.model_request].

    Supported by:

    * OpenAI
    * Anthropic (`'required'` and specific tools not supported with thinking enabled)
    * Google
    * Groq
    * Mistral
    * HuggingFace
    * Bedrock
    * xAI
    """

    seed: int
    """The random seed to use for the model, theoretically allowing for deterministic results.

    Supported by:

    * OpenAI
    * Groq
    * Cohere
    * Mistral
    * Gemini
    """

    presence_penalty: float
    """Penalize new tokens based on whether they have appeared in the text so far.

    Supported by:

    * OpenAI
    * Groq
    * Cohere
    * Gemini
    * Mistral
    * xAI
    """

    frequency_penalty: float
    """Penalize new tokens based on their existing frequency in the text so far.

    Supported by:

    * OpenAI
    * Groq
    * Cohere
    * Gemini
    * Mistral
    * xAI
    """

    logit_bias: dict[str, int]
    """Modify the likelihood of specified tokens appearing in the completion.

    Supported by:

    * OpenAI
    * Groq
    """

    stop_sequences: list[str]
    """Sequences that will cause the model to stop generating.

    Supported by:

    * OpenAI
    * Anthropic
    * Bedrock
    * Mistral
    * Groq
    * Cohere
    * Google
    * xAI
    """

    extra_headers: dict[str, str]
    """Extra headers to send to the model.

    Supported by:

    * OpenAI
    * Anthropic
    * Gemini
    * Groq
    * xAI
    """

    thinking: ThinkingLevel
    """Enable or configure thinking/reasoning for the model.

    - `True`: Enable thinking with the provider's default effort level.
    - `False`: Disable thinking (silently ignored if the model always thinks).
    - `'minimal'`/`'low'`/`'medium'`/`'high'`/`'xhigh'`: Enable thinking at a specific effort level.

    When omitted, the model uses its default behavior (which may include thinking
    for reasoning models).

    Provider-specific thinking settings (e.g., `anthropic_thinking`,
    `openai_reasoning_effort`) take precedence over this unified field.

    Supported by:

    * Anthropic
    * OpenAI
    * Gemini
    * Groq
    * Bedrock
    * OpenRouter
    * Cerebras
    * xAI
    """

    service_tier: ServiceTier
    """The cross-provider service tier to use for the model request.

    See [`ServiceTier`][pydantic_ai.settings.ServiceTier] for the value semantics and
    the per-provider mapping table. Provider-specific settings (`openai_service_tier`,
    `anthropic_service_tier`, `bedrock_service_tier`, `google_cloud_service_tier`)
    take precedence over this unified field when set.

    Supported by:

    * OpenAI
    * Anthropic
    * Bedrock
    * Google (Gemini API and Google Cloud)
    """

    extra_body: object
    """Extra body to send to the model.

    Supported by:

    * OpenAI
    * Anthropic
    * Groq
    """

max_tokens instance-attribute 

max_tokens: int

The maximum number of tokens to generate before stopping.

Supported by:

  • Gemini
  • Anthropic
  • OpenAI
  • Groq
  • Cohere
  • Mistral
  • Bedrock
  • MCP Sampling
  • xAI

temperature instance-attribute 

temperature: float

Amount of randomness injected into the response.

Use temperature closer to 0.0 for analytical / multiple choice, and closer to a model's maximum temperature for creative and generative tasks.

Note that even with temperature of 0.0, the results will not be fully deterministic.

Supported by:

  • Gemini
  • Anthropic
  • OpenAI
  • Groq
  • Cohere
  • Mistral
  • Bedrock
  • xAI

top_p instance-attribute 

top_p: float

An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass.

So 0.1 means only the tokens comprising the top 10% probability mass are considered.

You should either alter temperature or top_p, but not both.

Supported by:

  • Gemini
  • Anthropic
  • OpenAI
  • Groq
  • Cohere
  • Mistral
  • Bedrock
  • xAI

timeout instance-attribute 

timeout: float | Timeout

Override the client-level default timeout for a request, in seconds.

Supported by:

  • Gemini
  • Anthropic
  • OpenAI
  • Groq
  • Mistral
  • xAI

parallel_tool_calls instance-attribute 

parallel_tool_calls: bool

Whether to allow parallel tool calls.

Supported by:

  • OpenAI (some models, not o1)
  • Groq
  • Anthropic
  • xAI

tool_choice instance-attribute 

tool_choice: ToolChoice

Control which function tools the model can use.

See the Tool Choice guide for detailed documentation and examples.

  • None (default): Defaults to 'auto' behavior
  • 'auto': All tools available, model decides whether to use them
  • 'none': Disables function tools; model responds with text only (output tools remain for structured output)
  • 'required': Forces tool use; excludes output tools so the agent cannot produce a final response when set statically
  • list[str]: Only specified tools; excludes output tools so the agent cannot produce a final response when set statically
  • ToolOrOutput: Specified function tools plus output tools/text/image

Note: setting 'required' or list[str] statically (via the model_settings argument of [Agent.run][pydantic_ai.Agent.run] or the agent's own model_settings) raises a UserError, because it would force a tool call on every step and prevent the agent from producing a final response. To vary tool_choice per step (e.g. force a tool on the first step only), return a callable from a capability's get_model_settings — those values are trusted to adapt across steps. For single API calls without an agent loop, use pydantic_ai.direct.model_request.

Supported by:

  • OpenAI
  • Anthropic ('required' and specific tools not supported with thinking enabled)
  • Google
  • Groq
  • Mistral
  • HuggingFace
  • Bedrock
  • xAI

seed instance-attribute 

seed: int

The random seed to use for the model, theoretically allowing for deterministic results.

Supported by:

  • OpenAI
  • Groq
  • Cohere
  • Mistral
  • Gemini

presence_penalty instance-attribute 

presence_penalty: float

Penalize new tokens based on whether they have appeared in the text so far.

Supported by:

  • OpenAI
  • Groq
  • Cohere
  • Gemini
  • Mistral
  • xAI

frequency_penalty instance-attribute 

frequency_penalty: float

Penalize new tokens based on their existing frequency in the text so far.

Supported by:

  • OpenAI
  • Groq
  • Cohere
  • Gemini
  • Mistral
  • xAI

logit_bias instance-attribute 

logit_bias: dict[str, int]

Modify the likelihood of specified tokens appearing in the completion.

Supported by:

  • OpenAI
  • Groq

stop_sequences instance-attribute 

stop_sequences: list[str]

Sequences that will cause the model to stop generating.

Supported by:

  • OpenAI
  • Anthropic
  • Bedrock
  • Mistral
  • Groq
  • Cohere
  • Google
  • xAI

extra_headers instance-attribute 

extra_headers: dict[str, str]

Extra headers to send to the model.

Supported by:

  • OpenAI
  • Anthropic
  • Gemini
  • Groq
  • xAI

thinking instance-attribute 

thinking: ThinkingLevel

Enable or configure thinking/reasoning for the model.

  • True: Enable thinking with the provider's default effort level.
  • False: Disable thinking (silently ignored if the model always thinks).
  • 'minimal'/'low'/'medium'/'high'/'xhigh': Enable thinking at a specific effort level.

When omitted, the model uses its default behavior (which may include thinking for reasoning models).

Provider-specific thinking settings (e.g., anthropic_thinking, openai_reasoning_effort) take precedence over this unified field.

Supported by:

  • Anthropic
  • OpenAI
  • Gemini
  • Groq
  • Bedrock
  • OpenRouter
  • Cerebras
  • xAI

service_tier instance-attribute 

service_tier: ServiceTier

The cross-provider service tier to use for the model request.

See [ServiceTier][pydantic_ai.settings.ServiceTier] for the value semantics and the per-provider mapping table. Provider-specific settings (openai_service_tier, anthropic_service_tier, bedrock_service_tier, google_cloud_service_tier) take precedence over this unified field when set.

Supported by:

  • OpenAI
  • Anthropic
  • Bedrock
  • Google (Gemini API and Google Cloud)

extra_body instance-attribute 

extra_body: object

Extra body to send to the model.

Supported by:

  • OpenAI
  • Anthropic
  • Groq

ToolOrOutput dataclass

Restricts function tools while keeping output tools and direct text/image output available.

Use this when you want to control which function tools the model can use in an agent run while still allowing the agent to complete with structured output, text, or images.

See the Tool Choice guide for examples.

Source code in pydantic_ai_slim/pydantic_ai/settings.py 
27
28
29
30
31
32
33
34
35
36
37
38
39
@dataclass
class ToolOrOutput:
    """Restricts function tools while keeping output tools and direct text/image output available.

    Use this when you want to control which function tools the model can use
    in an agent run while still allowing the agent to complete with structured output,
    text, or images.

    See the [Tool Choice guide](../tools-advanced.md#tool-choice) for examples.
    """

    function_tools: list[str]
    """The names of function tools available to the model."""

function_tools instance-attribute 

function_tools: list[str]

The names of function tools available to the model.