pydantic-ai 0.7.5__tar.gz → 0.8.0__tar.gz

{pydantic_ai-0.7.5 → pydantic_ai-0.8.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pydantic-ai
-Version: 0.7.5
+Version: 0.8.0
 Summary: Agent Framework / shim to use Pydantic with LLMs
 Project-URL: Homepage, https://ai.pydantic.dev
 Project-URL: Source, https://github.com/pydantic/pydantic-ai
@@ -28,11 +28,11 @@ Classifier: Topic :: Internet
 Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Requires-Python: >=3.9
-Requires-Dist: pydantic-ai-slim[ag-ui,anthropic,bedrock,cli,cohere,evals,google,groq,huggingface,mcp,mistral,openai,retries,temporal,vertexai]==0.7.5
+Requires-Dist: pydantic-ai-slim[ag-ui,anthropic,bedrock,cli,cohere,evals,google,groq,huggingface,mcp,mistral,openai,retries,temporal,vertexai]==0.8.0
 Provides-Extra: a2a
 Requires-Dist: fasta2a>=0.4.1; extra == 'a2a'
 Provides-Extra: examples
-Requires-Dist: pydantic-ai-examples==0.7.5; extra == 'examples'
+Requires-Dist: pydantic-ai-examples==0.8.0; extra == 'examples'
 Provides-Extra: logfire
 Requires-Dist: logfire>=3.14.1; extra == 'logfire'
 Description-Content-Type: text/markdown

{pydantic_ai-0.7.5 → pydantic_ai-0.8.0}/tests/conftest.py

@@ -390,6 +390,11 @@ def heroku_inference_key() -> str:
     return os.getenv('HEROKU_INFERENCE_KEY', 'mock-api-key')
 
 
+@pytest.fixture(scope='session')
+def cerebras_api_key() -> str:
+    return os.getenv('CEREBRAS_API_KEY', 'mock-api-key')
+
+
 @pytest.fixture(scope='session')
 def bedrock_provider():
     try:
@@ -476,10 +481,10 @@ def model(
 
             return TestModel()
         elif request.param == 'openai':
-            from pydantic_ai.models.openai import OpenAIModel
+            from pydantic_ai.models.openai import OpenAIChatModel
             from pydantic_ai.providers.openai import OpenAIProvider
 
-            return OpenAIModel('o3-mini', provider=OpenAIProvider(api_key=openai_api_key))
+            return OpenAIChatModel('o3-mini', provider=OpenAIProvider(api_key=openai_api_key))
         elif request.param == 'anthropic':
             from pydantic_ai.models.anthropic import AnthropicModel
             from pydantic_ai.providers.anthropic import AnthropicProvider

{pydantic_ai-0.7.5 → pydantic_ai-0.8.0}/tests/json_body_serializer.py

@@ -69,6 +69,7 @@ def serialize(cassette_dict: Any):  # pragma: lax no cover
                     # Responses will have the body under a field called 'string'
                     body = body.get('string')
                 if body is not None:
+                    # NOTE(Marcelo): This doesn't handle gzip compression.
                     data['parsed_body'] = json.loads(body)  # pyright: ignore[reportUnknownArgumentType]
                     if 'access_token' in data['parsed_body']:
                         data['parsed_body']['access_token'] = 'scrubbed'

{pydantic_ai-0.7.5 → pydantic_ai-0.8.0}/tests/mcp_server.py

@@ -3,8 +3,7 @@ from pathlib import Path
 from typing import Any
 
 from mcp.server.fastmcp import Context, FastMCP, Image
-from mcp.server.session import ServerSessionT
-from mcp.shared.context import LifespanContextT, RequestT
+from mcp.server.session import ServerSession
 from mcp.types import (
     BlobResourceContents,
     EmbeddedResource,
@@ -13,7 +12,7 @@ from mcp.types import (
     TextContent,
     TextResourceContents,
 )
-from pydantic import AnyUrl
+from pydantic import AnyUrl, BaseModel
 
 mcp = FastMCP('Pydantic AI MCP Server')
 log_level = 'unset'
@@ -170,7 +169,7 @@ async def get_log_level(ctx: Context) -> str:  # type: ignore
 
 
 @mcp.tool()
-async def echo_deps(ctx: Context[ServerSessionT, LifespanContextT, RequestT]) -> dict[str, Any]:
+async def echo_deps(ctx: Context[ServerSession, None]) -> dict[str, Any]:
     """Echo the run context.
 
     Args:
@@ -186,7 +185,7 @@ async def echo_deps(ctx: Context[ServerSessionT, LifespanContextT, RequestT]) ->
 
 
 @mcp.tool()
-async def use_sampling(ctx: Context, foo: str) -> str:  # type: ignore
+async def use_sampling(ctx: Context[ServerSession, None], foo: str) -> str:
     """Use sampling callback."""
 
     result = await ctx.session.create_message(
@@ -202,6 +201,22 @@ async def use_sampling(ctx: Context, foo: str) -> str:  # type: ignore
     return result.model_dump_json(indent=2)
 
 
+class UserResponse(BaseModel):
+    response: str
+
+
+@mcp.tool()
+async def use_elicitation(ctx: Context[ServerSession, None], question: str) -> str:
+    """Use elicitation callback to ask the user a question."""
+
+    result = await ctx.elicit(message=question, schema=UserResponse)
+
+    if result.action == 'accept' and result.data:
+        return f'User responded: {result.data.response}'
+    else:
+        return f'User {result.action}ed the elicitation'
+
+
 @mcp._mcp_server.set_logging_level()  # pyright: ignore[reportPrivateUsage]
 async def set_logging_level(level: str) -> None:
     global log_level

{pydantic_ai-0.7.5 → pydantic_ai-0.8.0}/tests/models/cassettes/test_model_names/test_known_model_names.yaml

@@ -83,4 +83,69 @@ interactions:
     status:
       code: 200
       message: OK
+- request:
+    body: ''
+    headers:
+      accept:
+      - application/json
+      connection:
+      - keep-alive
+      host:
+      - api.cerebras.ai
+    method: GET
+    uri: https://api.cerebras.ai/v1/models
+  response:
+    headers:
+      alt-svc:
+      - h3=":443"; ma=86400
+      content-length:
+      - '762'
+      content-type:
+      - application/json
+      referrer-policy:
+      - strict-origin-when-cross-origin
+      strict-transport-security:
+      - max-age=3600; includeSubDomains
+    parsed_body:
+      data:
+      - created: 0
+        id: llama-4-maverick-17b-128e-instruct
+        object: model
+        owned_by: Cerebras
+      - created: 0
+        id: qwen-3-32b
+        object: model
+        owned_by: Cerebras
+      - created: 0
+        id: qwen-3-235b-a22b-instruct-2507
+        object: model
+        owned_by: Cerebras
+      - created: 0
+        id: llama-4-scout-17b-16e-instruct
+        object: model
+        owned_by: Cerebras
+      - created: 0
+        id: gpt-oss-120b
+        object: model
+        owned_by: Cerebras
+      - created: 0
+        id: qwen-3-coder-480b
+        object: model
+        owned_by: Cerebras
+      - created: 0
+        id: llama-3.3-70b
+        object: model
+        owned_by: Cerebras
+      - created: 0
+        id: llama3.1-8b
+        object: model
+        owned_by: Cerebras
+      - created: 0
+        id: qwen-3-235b-a22b-thinking-2507
+        object: model
+        owned_by: Cerebras
+      object: list
+    status:
+      code: 200
+      message: OK
 version: 1

pydantic_ai-0.8.0/tests/models/cassettes/test_openai/test_openai_model_cerebras_provider.yaml

@@ -0,0 +1,69 @@
+interactions:
+- request:
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '111'
+      content-type:
+      - application/json
+      host:
+      - api.cerebras.ai
+    method: POST
+    parsed_body:
+      messages:
+      - content: What is the capital of France?
+        role: user
+      model: llama3.3-70b
+      stream: false
+    uri: https://api.cerebras.ai/v1/chat/completions
+  response:
+    headers:
+      alt-svc:
+      - h3=":443"; ma=86400
+      connection:
+      - keep-alive
+      content-length:
+      - '558'
+      content-type:
+      - application/json
+      inference-id:
+      - chatcmpl-175bff47-ce89-4fad-bead-62a14f787c49
+      referrer-policy:
+      - strict-origin-when-cross-origin
+      strict-transport-security:
+      - max-age=3600; includeSubDomains
+      transfer-encoding:
+      - chunked
+    parsed_body:
+      choices:
+      - finish_reason: stop
+        index: 0
+        message:
+          content: The capital of France is Paris.
+          role: assistant
+      created: 1756198141
+      id: chatcmpl-175bff47-ce89-4fad-bead-62a14f787c49
+      model: llama-3.3-70b
+      object: chat.completion
+      system_fingerprint: fp_b6a51bb1d26f03ea3772
+      time_info:
+        completion_time: 0.003633212
+        created: 1756198141
+        prompt_time: 0.002251218
+        queue_time: 0.000199151
+        total_time: 0.007346153259277344
+      usage:
+        completion_tokens: 8
+        prompt_tokens: 42
+        prompt_tokens_details:
+          cached_tokens: 0
+        total_tokens: 50
+    status:
+      code: 200
+      message: OK
+version: 1

pydantic_ai-0.8.0/tests/models/cassettes/test_openai/test_openai_model_cerebras_provider_harmony.yaml

@@ -0,0 +1,70 @@
+interactions:
+- request:
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '111'
+      content-type:
+      - application/json
+      host:
+      - api.cerebras.ai
+    method: POST
+    parsed_body:
+      messages:
+      - content: What is the capital of France?
+        role: user
+      model: gpt-oss-120b
+      stream: false
+    uri: https://api.cerebras.ai/v1/chat/completions
+  response:
+    headers:
+      alt-svc:
+      - h3=":443"; ma=86400
+      connection:
+      - keep-alive
+      content-length:
+      - '661'
+      content-type:
+      - application/json
+      inference-id:
+      - chatcmpl-bc3bbd04-e8df-4ab6-bd82-9fb33726cb93
+      referrer-policy:
+      - strict-origin-when-cross-origin
+      strict-transport-security:
+      - max-age=3600; includeSubDomains
+      transfer-encoding:
+      - chunked
+    parsed_body:
+      choices:
+      - finish_reason: stop
+        index: 0
+        message:
+          content: The capital of France is **Paris**.
+          reasoning: 'User asks simple question: capital of France. Answer: Paris. Provide concise answer.'
+          role: assistant
+      created: 1756199010
+      id: chatcmpl-bc3bbd04-e8df-4ab6-bd82-9fb33726cb93
+      model: gpt-oss-120b
+      object: chat.completion
+      system_fingerprint: fp_a5fe16f102a24cf8caf1
+      time_info:
+        completion_time: 0.026440676
+        created: 1756199010
+        prompt_time: 0.002999482
+        queue_time: 0.000899766
+        total_time: 0.03260970115661621
+      usage:
+        completion_tokens: 36
+        prompt_tokens: 74
+        prompt_tokens_details:
+          cached_tokens: 0
+        total_tokens: 110
+    status:
+      code: 200
+      message: OK
+version: 1

pydantic_ai-0.8.0/tests/models/cassettes/test_openai/test_openai_model_cerebras_provider_qwen_3_coder.yaml

@@ -0,0 +1,189 @@
+interactions:
+- request:
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '414'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=hqG7YTrLAdrCNKQU34H3qhIKxESTvFsG9X7OjEzKuDk-1756198142-1.0.1.1-WG8.rT3zC66xxHN6c33AcERCdH0FDum2a7djbEu85gy9rEiq9nTHQVY8g5BAZ_DmviuVxhphMwTsxDclYqQCobIlgurA9g.prg5M_i9PZyM
+      host:
+      - api.cerebras.ai
+    method: POST
+    parsed_body:
+      messages:
+      - content: What is the capital of France?
+        role: user
+      model: qwen-3-coder-480b
+      stream: false
+      tool_choice: auto
+      tools:
+      - function:
+          description: The final response which ends this conversation
+          name: final_result
+          parameters:
+            properties:
+              city:
+                type: string
+              country:
+                type: string
+            required:
+            - city
+            - country
+            title: Location
+            type: object
+        type: function
+    uri: https://api.cerebras.ai/v1/chat/completions
+  response:
+    headers:
+      alt-svc:
+      - h3=":443"; ma=86400
+      connection:
+      - keep-alive
+      content-length:
+      - '645'
+      content-type:
+      - application/json
+      inference-id:
+      - chatcmpl-65763e70-ba85-41a3-a758-4dc3fcb73c9a
+      referrer-policy:
+      - strict-origin-when-cross-origin
+      strict-transport-security:
+      - max-age=3600; includeSubDomains
+      transfer-encoding:
+      - chunked
+    parsed_body:
+      choices:
+      - finish_reason: stop
+        index: 0
+        message:
+          content: The capital of France is Paris. If you need more information about Paris or any other details, feel free
+            to ask!
+          role: assistant
+      created: 1756198142
+      id: chatcmpl-65763e70-ba85-41a3-a758-4dc3fcb73c9a
+      model: qwen-3-coder-480b
+      object: chat.completion
+      system_fingerprint: fp_4405be376d055dc5bff7
+      time_info:
+        completion_time: 0.045860486
+        created: 1756198142
+        prompt_time: 0.011004894
+        queue_time: 0.001344637
+        total_time: 0.05971097946166992
+      usage:
+        completion_tokens: 25
+        prompt_tokens: 304
+        prompt_tokens_details:
+          cached_tokens: 0
+        total_tokens: 329
+    status:
+      code: 200
+      message: OK
+- request:
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '727'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=hqG7YTrLAdrCNKQU34H3qhIKxESTvFsG9X7OjEzKuDk-1756198142-1.0.1.1-WG8.rT3zC66xxHN6c33AcERCdH0FDum2a7djbEu85gy9rEiq9nTHQVY8g5BAZ_DmviuVxhphMwTsxDclYqQCobIlgurA9g.prg5M_i9PZyM
+      host:
+      - api.cerebras.ai
+    method: POST
+    parsed_body:
+      messages:
+      - content: What is the capital of France?
+        role: user
+      - content: The capital of France is Paris. If you need more information about Paris or any other details, feel free
+          to ask!
+        role: assistant
+      - content: |-
+          Validation feedback:
+          Plain text responses are not permitted, please include your response in a tool call
+
+          Fix the errors and try again.
+        role: user
+      model: qwen-3-coder-480b
+      stream: false
+      tool_choice: auto
+      tools:
+      - function:
+          description: The final response which ends this conversation
+          name: final_result
+          parameters:
+            properties:
+              city:
+                type: string
+              country:
+                type: string
+            required:
+            - city
+            - country
+            title: Location
+            type: object
+        type: function
+    uri: https://api.cerebras.ai/v1/chat/completions
+  response:
+    headers:
+      alt-svc:
+      - h3=":443"; ma=86400
+      connection:
+      - keep-alive
+      content-length:
+      - '672'
+      content-type:
+      - application/json
+      inference-id:
+      - chatcmpl-6659125c-6d06-402c-be3b-4640aa64ffa2
+      referrer-policy:
+      - strict-origin-when-cross-origin
+      strict-transport-security:
+      - max-age=3600; includeSubDomains
+      transfer-encoding:
+      - chunked
+    parsed_body:
+      choices:
+      - finish_reason: tool_calls
+        index: 0
+        message:
+          role: assistant
+          tool_calls:
+          - function:
+              arguments: '{"city": "Paris", "country": "France"}'
+              name: final_result
+            id: b8847f144
+            type: function
+      created: 1756198143
+      id: chatcmpl-6659125c-6d06-402c-be3b-4640aa64ffa2
+      model: qwen-3-coder-480b
+      object: chat.completion
+      system_fingerprint: fp_4405be376d055dc5bff7
+      time_info:
+        completion_time: 0.103362557
+        created: 1756198143
+        prompt_time: 0.22219986
+        queue_time: 0.346436935
+        total_time: 0.6736249923706055
+      usage:
+        completion_tokens: 33
+        prompt_tokens: 364
+        prompt_tokens_details:
+          cached_tokens: 0
+        total_tokens: 397
+    status:
+      code: 200
+      message: OK
+version: 1

pydantic_ai-0.8.0/tests/models/cassettes/test_openai_responses/test_openai_responses_usage_without_tokens_details.yaml

@@ -0,0 +1,96 @@
+interactions:
+- request:
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '84'
+      content-type:
+      - application/json
+      host:
+      - api.openai.com
+    method: POST
+    parsed_body:
+      input:
+      - content: What is 2+2?
+        role: user
+      model: gpt-4o
+      stream: false
+    uri: https://api.openai.com/v1/responses
+  response:
+    headers:
+      alt-svc:
+      - h3=":443"; ma=86400
+      connection:
+      - keep-alive
+      content-length:
+      - '1369'
+      content-type:
+      - application/json
+      openai-organization:
+      - pydantic-28gund
+      openai-processing-ms:
+      - '899'
+      openai-project:
+      - proj_dKobscVY9YJxeEaDJen54e3d
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      transfer-encoding:
+      - chunked
+    parsed_body:
+      background: false
+      created_at: 1756246493
+      error: null
+      id: resp_68ae31dd0edc819da9e77ad3cdb128770306cb62070aed80
+      incomplete_details: null
+      instructions: null
+      max_output_tokens: null
+      max_tool_calls: null
+      metadata: {}
+      model: gpt-4o-2024-08-06
+      object: response
+      output:
+      - content:
+        - annotations: []
+          logprobs: []
+          text: 2 + 2 equals 4.
+          type: output_text
+        id: msg_68ae31ddcef0819da13eb00d2b393eb60306cb62070aed80
+        role: assistant
+        status: completed
+        type: message
+      parallel_tool_calls: true
+      previous_response_id: null
+      prompt_cache_key: null
+      reasoning:
+        effort: null
+        summary: null
+      safety_identifier: null
+      service_tier: default
+      status: completed
+      store: true
+      temperature: 1.0
+      text:
+        format:
+          type: text
+        verbosity: medium
+      tool_choice: auto
+      tools: []
+      top_logprobs: 0
+      top_p: 1.0
+      truncation: disabled
+      usage:
+        input_tokens: 14
+        output_tokens: 9
+        total_tokens: 23
+      user: null
+    status:
+      code: 200
+      message: OK
+version: 1

{pydantic_ai-0.7.5 → pydantic_ai-0.8.0}/tests/models/test_deepseek.py

@@ -24,7 +24,7 @@ from pydantic_ai.usage import RequestUsage
 from ..conftest import IsDatetime, IsStr, try_import
 
 with try_import() as imports_successful:
-    from pydantic_ai.models.openai import OpenAIModel
+    from pydantic_ai.models.openai import OpenAIChatModel
     from pydantic_ai.providers.deepseek import DeepSeekProvider
 
 
@@ -36,7 +36,7 @@ pytestmark = [
 
 
 async def test_deepseek_model_thinking_part(allow_model_requests: None, deepseek_api_key: str):
-    deepseek_model = OpenAIModel('deepseek-reasoner', provider=DeepSeekProvider(api_key=deepseek_api_key))
+    deepseek_model = OpenAIChatModel('deepseek-reasoner', provider=DeepSeekProvider(api_key=deepseek_api_key))
     agent = Agent(model=deepseek_model)
     result = await agent.run('How do I cross the street?')
     assert result.all_messages() == snapshot(
@@ -63,7 +63,7 @@ async def test_deepseek_model_thinking_part(allow_model_requests: None, deepseek
 
 
 async def test_deepseek_model_thinking_stream(allow_model_requests: None, deepseek_api_key: str):
-    deepseek_model = OpenAIModel('deepseek-reasoner', provider=DeepSeekProvider(api_key=deepseek_api_key))
+    deepseek_model = OpenAIChatModel('deepseek-reasoner', provider=DeepSeekProvider(api_key=deepseek_api_key))
     agent = Agent(model=deepseek_model)
 
     event_parts: list[Any] = []

{pydantic_ai-0.7.5 → pydantic_ai-0.8.0}/tests/models/test_model.py

@@ -10,10 +10,10 @@ from ..conftest import TestEnv
 # TODO(Marcelo): We need to add Vertex AI to the test cases.
 
 TEST_CASES = [
-    ('OPENAI_API_KEY', 'openai:gpt-3.5-turbo', 'gpt-3.5-turbo', 'openai', 'openai', 'OpenAIModel'),
-    ('OPENAI_API_KEY', 'gpt-3.5-turbo', 'gpt-3.5-turbo', 'openai', 'openai', 'OpenAIModel'),
-    ('OPENAI_API_KEY', 'o1', 'o1', 'openai', 'openai', 'OpenAIModel'),
-    ('AZURE_OPENAI_API_KEY', 'azure:gpt-3.5-turbo', 'gpt-3.5-turbo', 'azure', 'azure', 'OpenAIModel'),
+    ('OPENAI_API_KEY', 'openai:gpt-3.5-turbo', 'gpt-3.5-turbo', 'openai', 'openai', 'OpenAIChatModel'),
+    ('OPENAI_API_KEY', 'gpt-3.5-turbo', 'gpt-3.5-turbo', 'openai', 'openai', 'OpenAIChatModel'),
+    ('OPENAI_API_KEY', 'o1', 'o1', 'openai', 'openai', 'OpenAIChatModel'),
+    ('AZURE_OPENAI_API_KEY', 'azure:gpt-3.5-turbo', 'gpt-3.5-turbo', 'azure', 'azure', 'OpenAIChatModel'),
     ('GEMINI_API_KEY', 'google-gla:gemini-1.5-flash', 'gemini-1.5-flash', 'google-gla', 'google', 'GoogleModel'),
     ('GEMINI_API_KEY', 'gemini-1.5-flash', 'gemini-1.5-flash', 'google-gla', 'google', 'GoogleModel'),
     (
@@ -70,7 +70,7 @@ TEST_CASES = [
         'xai/grok-3-mini',
         'github',
         'github',
-        'OpenAIModel',
+        'OpenAIChatModel',
     ),
     (
         'MOONSHOTAI_API_KEY',
@@ -78,7 +78,7 @@ TEST_CASES = [
         'kimi-k2-0711-preview',
         'moonshotai',
         'moonshotai',
-        'OpenAIModel',
+        'OpenAIChatModel',
     ),
     (
         'GROK_API_KEY',
@@ -86,7 +86,7 @@ TEST_CASES = [
         'grok-3',
         'grok',
         'grok',
-        'OpenAIModel',
+        'OpenAIChatModel',
     ),
     (
         'GROK_API_KEY',
@@ -94,7 +94,7 @@ TEST_CASES = [
         'grok-4',
         'grok',
         'grok',
-        'OpenAIModel',
+        'OpenAIChatModel',
     ),
     (
         'OPENAI_API_KEY',