Overview

15 Releases published by 3 people

• v0.5.0 v0.5.0 (2025年08月04日)

  published Aug 4, 2025

• v0.5.1 v0.5.1 (2025年08月06日)

  published Aug 6, 2025

• v0.6.0 v0.6.0 (2025年08月06日)

  published Aug 6, 2025

• v0.6.1 v0.6.1 (2025年08月07日)

  published Aug 7, 2025

• v0.6.2 v0.6.2 (2025年08月07日)

  published Aug 7, 2025

• v0.7.0 v0.7.0 (2025年08月12日)

  published Aug 12, 2025

• v0.7.1 v0.7.1 (2025年08月13日)

  published Aug 13, 2025

• v0.7.2 v0.7.2 (2025年08月14日)

  published Aug 14, 2025

• v0.7.3 v0.7.3 (2025年08月19日)

  published Aug 19, 2025

• v0.7.4 v0.7.4 (2025年08月20日)

  published Aug 20, 2025

• v0.7.5 v0.7.5 (2025年08月25日)

  published Aug 25, 2025

• v0.7.6 v0.7.6 (2025年08月26日)

  published Aug 26, 2025

• v0.8.0 v0.8.0 (2025年08月26日)

  published Aug 26, 2025

• v0.8.1 v0.8.1 (2025年08月29日)

  published Aug 29, 2025

• v1.0.0b1 v1.0.0b1 (2025年08月30日)

  published Aug 30, 2025

172 Pull requests merged by 41 people

• Raise error when trying to use Google built-in tools with user/output tools

  #2777 merged Sep 2, 2025

• Handle Groq tool_use_failed errors by getting model to retry

  #2774 merged Sep 2, 2025

• Create separate ThinkingParts for separate OpenAI Responses reasoning summary parts

  #2775 merged Sep 2, 2025

• Support NativeOutput with Groq

  #2772 merged Sep 2, 2025

• Fix Anthropic streaming usage counting

  #2771 merged Sep 2, 2025

• Reorganize the tools docs

  #2749 merged Sep 2, 2025

• docs: Add LangWatch to list of observability platforms

  #2680 merged Sep 2, 2025

• add identifier field support to FileUrl and subclasses

  #2636 merged Sep 2, 2025

• Minor fixes: Agent.__aenter__ returns Self, use default instrumentation for MCP sampling model

  #2765 merged Sep 2, 2025

• Properly close VertexAI HTTP client at end of live test

  #2768 merged Sep 2, 2025

• Include project ID for Google service account auth

  #2755 merged Sep 1, 2025

• Reorganize the docs nav hierarchy

  #2748 merged Sep 1, 2025

• Close original response when retrying HTTP request

  #2753 merged Sep 1, 2025

• fix(examples): Check for BaseModel before serializing json

  #2729 merged Sep 1, 2025

• Fix error when streaming from Gemini includes only executable_code or code_execution_result

  #2719 merged Sep 1, 2025

• Add docstring_format, require_parameter_descriptions, schema_generator to FunctionToolset

  #2601 merged Sep 1, 2025

• Raise error if MCP server __aexit__ is called when _running_count is already 0

  #2696 merged Sep 1, 2025

• Update mcp server package version

  #2741 merged Sep 1, 2025

• Don't emit empty AG-UI thinking message events

  #2754 merged Sep 1, 2025

• Fix HITL tool approval docs example

  #2752 merged Sep 1, 2025

• Add beta release admonition

  #2744 merged Aug 30, 2025

• Fix newsletter signup form

  #2743 merged Aug 30, 2025

• Remove more todo comments

  #2739 merged Aug 30, 2025

• BREAKING CHANGE: Make many more dataclasses kw-only

  #2738 merged Aug 30, 2025

• V1: Remove cases and averages from eval span

  #2715 merged Aug 29, 2025

• Default InstrumentationSettings version to 2

  #2726 merged Aug 29, 2025

• Let almost all types used in docs examples be imported directly from pydantic_ai

  #2736 merged Aug 29, 2025

• V1: Remove errors when passing Retrying instead of RetryConfig to TenacityTransport

  #2717 merged Aug 29, 2025

• V1: Remove Temporal docs beta header

  #2727 merged Aug 29, 2025

• Restore DeferredToolCalls with deprecation warning

  #2735 merged Aug 29, 2025

• Include logfire with pydantic-ai package

  #2683 merged Aug 29, 2025

• Gracefully handle errors in evals

  #2295 merged Aug 29, 2025

• Add support for human-in-the-loop tool call approval

  #2581 merged Aug 29, 2025

• Deprecate OpenAIModelProfile.openai_supports_sampling_settings

  #2730 merged Aug 29, 2025

• Remove 'only on 3.10+' callouts

  #2733 merged Aug 29, 2025

• Use match statement now that we're dropping 3.9

  #2732 merged Aug 29, 2025

• Remove/update some TODO comments

  #2731 merged Aug 29, 2025

• Drop support for Python 3.9

  #2725 merged Aug 29, 2025

• Resolve TODO comment

  #2728 merged Aug 29, 2025

• V1: Rename ModelResponse.provider_request_id to provider_response_id

  #2710 merged Aug 29, 2025

• V1: Deprecate specifying a model name without a provider prefix, and the vertexai provider name

  #2711 merged Aug 28, 2025

• Remove unused non-user-code-prefixes stuff

  #2712 merged Aug 28, 2025

• Reenable vertexai docs test

  #2709 merged Aug 28, 2025

• ci: Fix pydantic-ai-slim job skipping all tests

  #2704 merged Aug 28, 2025

• Bump temporalio to 1.16.0

  #2703 merged Aug 27, 2025

• Add MockOpenAIResponses for use in tests

  #2702 merged Aug 27, 2025

• docs: Add message_history parameter documentation for CLI methods

  #2695 merged Aug 27, 2025

• Rename StreamedRunResult methods to be consistent with AgentStream

  #2692 merged Aug 27, 2025

• Add gen_ai.system_instructions attribute to agent run spans

  #2699 merged Aug 27, 2025

• Add v0.8.0 breaking changes to changelog

  #2694 merged Aug 26, 2025

• fix: Make FileUrl.media_type work correctly with inline-snapshot

  #2693 merged Aug 26, 2025

• fix: Handle missing token details in vLLM/OpenAI-compatible APIs

  #2669 merged Aug 26, 2025

• BREAKING CHANGE: Make AgentStreamEvent union of ModelResponseStreamEvent and HandleResponseEvent

  #2689 merged Aug 26, 2025

• docs: Add docstrings for EndStrategy, capture_run_messages, HandleResponseEvent to API docs

  #2688 merged Aug 26, 2025

• Add message_history parameter to agent.to_cli()

  #2674 merged Aug 26, 2025

• Fix docs for MCP elicitation callback

  #2685 merged Aug 26, 2025

• fix: Properly deserialize complex tool arguments with Temporal

  #2686 merged Aug 26, 2025

• Fix serialization / deserialization of FileUrl.media_type

  #2677 merged Aug 26, 2025

• feat: add elicitation callback support to MCP servers

  #2373 merged Aug 26, 2025

• Add Cerebras provider

  #2643 merged Aug 26, 2025

• BREAKING CHANGE: Fix tenacity implementation for improved retry behavior

  #2670 merged Aug 26, 2025

• fix(bedrock): skip SystemPromptPart with empty content

  #2672 merged Aug 26, 2025

• anthropic: drop new lines on empty system prompt

  #2678 merged Aug 26, 2025

• Deprecate OpenAIModel in favor of OpenAIChatModel

  #2676 merged Aug 26, 2025

• Add newsletter form to docs

  #2673 merged Aug 26, 2025

• Fix inability to call response.raise_for_status in AsyncTenacityTransport

  #2668 merged Aug 26, 2025

• Replace all_messages_events with pydantic_ai.all_messages for InstrumentationSettings(version=2)

  #2658 merged Aug 25, 2025

• Fix multi-agent example token limit in documentation

  #2662 merged Aug 25, 2025

• Ensure content is always set for assistant tool call messages for OpenAI.

  #2641 merged Aug 25, 2025

• fix coverage badge

  #2652 merged Aug 24, 2025

• Add AGENTS.md

  #2637 merged Aug 21, 2025

• Use new OpenTelemetry GenAI chat span attribute conventions

  #2349 merged Aug 21, 2025

• Allow proper type on AnthropicProvider when using Bedrock

  #2490 merged Aug 21, 2025

• Add span_id and trace_id to EvaluationReport

  #2627 merged Aug 21, 2025

• Include thoughts tokens in output_tokens for Google models

  #2634 merged Aug 21, 2025

• Add price() method to ModelResponse

  #2584 merged Aug 21, 2025

• Handle 'STOP' finish_reason in GeminiStreamedResponse

  #2631 merged Aug 21, 2025

• Fix tests for google-genai>=1.31.0

  #2625 merged Aug 21, 2025

• docs: indicate Google model support for WebSearchTool

  #2624 merged Aug 20, 2025

• Drop assertion on Google streaming

  #2618 merged Aug 20, 2025

• Update mcp-run-python docs

  #2610 merged Aug 20, 2025

• Add missing UrlContextTool into __all__

  #2617 merged Aug 20, 2025

• feat: support Google's url_context builtin tool

  #2604 merged Aug 20, 2025

• Add takes_ctx arg to Tool.from_schema

  #2615 merged Aug 20, 2025

• Fix bug with google model safety handling

  #2066 merged Aug 19, 2025

• Use _provider.name instead of _system

  #2596 merged Aug 18, 2025

• Fix typo in model argument documentation

  #2579 merged Aug 18, 2025

• tests: improve the time for some tests

  #2592 merged Aug 18, 2025

• Rename config.yaml to config.yml

  #2585 merged Aug 16, 2025

• tests: drop commonly modified headers from VCR

  #2583 merged Aug 16, 2025

• Pin temporalio to 1.15.0 as plugins API is still experimental

  #2582 merged Aug 15, 2025

• Clarify usage limits priority in docs

  #2576 merged Aug 15, 2025

• Add /cp command to CLI to copy last response to clipboard

  #2386 merged Aug 15, 2025

• Move system_prompt_role from OpenAIModel to OpenAIModelProfile

  #2573 merged Aug 15, 2025

• Make FallbackModel accept string model names

  #2564 merged Aug 15, 2025

• Deprecate Usage in favour of RequestUsage and RunUsage

  #2378 merged Aug 15, 2025

• Let message history end on ModelResponse and execute pending tool calls

  #2562 merged Aug 14, 2025

• Fix: Forward max_uses parameter to Anthropic WebSearchTool

  #2561 merged Aug 14, 2025

• chore: Improve Temporal test to check trace as tree instead of list

  #2559 merged Aug 14, 2025

• Ask model to try again if it produced a response without text or tool calls, only thinking

  #2556 merged Aug 14, 2025

• Ignore leading whitespace when streaming from Qwen or DeepSeek

  #2554 merged Aug 14, 2025

• chore: Fix path to agent class in CLAUDE.md

  #2550 merged Aug 14, 2025

• docs: Clarify why AG-UI example links are on localhost

  #2549 merged Aug 14, 2025

• Remove anthropic-beta default header set in AnthropicModel

  #2544 merged Aug 14, 2025

• Only wrap necessary parts of type aliases in forward annotations

  #2548 merged Aug 14, 2025

• Disable instrumentation by default in tests

  #2535 merged Aug 14, 2025

• Stop calling MCP server get_tools ahead of agent run span

  #2545 merged Aug 13, 2025

• chore: Fix uv.lock

  #2546 merged Aug 13, 2025

• Add UsageLimits.count_tokens_before_request using Gemini count_tokens API

  #2137 merged Aug 13, 2025

• Add support for OpenAI verbosity parameter in Responses API

  #2493 merged Aug 13, 2025

• Fix FallbackModel to respect each model's model settings

  #2540 merged Aug 13, 2025

• Add new OpenAI GPT-5 models

  #2503 merged Aug 13, 2025

• Make some minor tweaks to the temporal docs

  #2522 merged Aug 13, 2025

• Cleanup snapshot in test_evaluate_async_logfire

  #2538 merged Aug 13, 2025

• Add --durations=100 to all pytest calls in CI

  #2534 merged Aug 13, 2025

• Add genai-prices to dev deps and a basic test

  #2537 merged Aug 13, 2025

• Add pip to dev deps for PyCharm

  #2533 merged Aug 13, 2025

• ci: improve matrix readability

  #2532 merged Aug 13, 2025

• chore: drop hyperlint

  #2531 merged Aug 13, 2025

• chore: mention what to do with the documentation when deprecating a class

  #2530 merged Aug 13, 2025

• [docs]: Confident AI

  #2529 merged Aug 13, 2025

• Add support for "openai-responses" model inference string

  #2528 merged Aug 13, 2025

• Add best practices regarding to coverage to CLAUDE.md

  #2527 merged Aug 13, 2025

• Improve CLAUDE.md

  #2524 merged Aug 13, 2025

• Install only dev in CI

  #2523 merged Aug 13, 2025

• Let Agent be run in a Temporal workflow by moving model requests, tool calls, and MCP to Temporal activities

  #2225 merged Aug 12, 2025

• Make OpenAIResponsesModelSettings.openai_builtin_tools work again

  #2520 merged Aug 12, 2025

• Fix KeyError when parsing video metadata without audio track in Google models

  #2507 merged Aug 12, 2025

• Skip testing flaky evals example

  #2518 merged Aug 12, 2025

• Update config.yaml

  #2514 merged Aug 12, 2025

• ci: use the right install command

  #2506 merged Aug 11, 2025

• ci: split test suite

  #2436 merged Aug 11, 2025

• History processor replaces message history

  #2324 merged Aug 11, 2025

• fix: Add gpt-5 models to reasoning model detection for temperature parameter handling

  #2483 merged Aug 11, 2025

• ci: remove html and xml reports

  #2491 merged Aug 11, 2025

• tests: change time of evals example

  #2501 merged Aug 11, 2025

• Add ToolsetFunc to API docs (fix CI)

  #2486 merged Aug 8, 2025

• Let toolsets be built dynamically based on run context

  #2366 merged Aug 8, 2025

• Add AbstractAgent, WrapperAgent, Agent.event_stream_handler, Toolset.id, Agent.override(tools=...) in preparation for Temporal

  #2458 merged Aug 8, 2025

• Remove fail_under from top-level coverage config so <100% html-coverage step doesn't end CI run

  #2475 merged Aug 7, 2025

• Bump mcp-run-python

  #2470 merged Aug 7, 2025

• Add builtin_tools to Agent

  #2102 merged Aug 7, 2025

• Add new Heroku models

  #2459 merged Aug 7, 2025

• Support AWS_BEARER_TOKEN_BEDROCK API key env var

  #2456 merged Aug 6, 2025

• Send ThinkingParts back to Anthropic used through Bedrock

  #2454 merged Aug 6, 2025

• Make InlineDefsJsonSchemaTransformer public

  #2455 merged Aug 6, 2025

• Automatically use OpenAI strict mode for strict-compatible native output types

  #2447 merged Aug 6, 2025

• BREAKING CHANGE: drop 4 months old deprecation warnings

  #2451 merged Aug 6, 2025

• Deprecate GoogleGLAProvider and GoogleVertexProvider

  #2450 merged Aug 6, 2025

• BREAKING CHANGE: Remove result_type parameter and similar from Agent

  #2441 merged Aug 6, 2025

• BREAKING CHANGE: Remove format_as_xml module

  #2446 merged Aug 6, 2025

• docs: add griffe_warnings_deprecated

  #2444 merged Aug 6, 2025

• BREAKING CHANGE: Remove get_data and validate_structured_result from StreamedRunResult

  #2445 merged Aug 6, 2025

• BREAKING CHANGE: Remove data from FinalResult

  #2443 merged Aug 6, 2025

• BREAKING CHANGE: Remove next() method from Graph

  #2440 merged Aug 6, 2025

• Remove older deprecated models and add new model of Anthropic

  #2435 merged Aug 6, 2025

• Use httpx on GoogleProvider

  #2438 merged Aug 6, 2025

• Deprecate GeminiModel in favor of GoogleModel

  #2416 merged Aug 6, 2025

• typo in output docs

  #2427 merged Aug 6, 2025

• google: add more information about schema on union

  #2426 merged Aug 5, 2025

• tests: speed up the test suite

  #2414 merged Aug 5, 2025

• Fix tests for Logfire>=3.22.0

  #2346 merged Aug 5, 2025

• Document that tools become toolset internally

  #2395 merged Aug 4, 2025

• Document performance implications of async vs sync tools

  #2298 merged Aug 4, 2025

• BREAKING CHANGE: Support printing reasons in the console output for pydantic-evals

  #2163 merged Aug 4, 2025

• Fix ImageUrl, VideoUrl, AudioUrl and DocumentUrl not being serializable

  #2422 merged Aug 4, 2025

• BREAKING CHANGE: Change type of 'source' field on EvaluationResult

  #2388 merged Aug 4, 2025

• Let more BaseModels use OpenAI strict JSON mode by defaulting to additionalProperties=False

  #2419 merged Aug 4, 2025

• Allow string format, pattern and others in OpenAI strict JSON mode

  #2420 merged Aug 4, 2025

• Fix test_download_item_no_content_type test failing on macOS

  #2404 merged Aug 4, 2025

• Include default values in tool arguments JSON schema

  #2418 merged Aug 4, 2025

• docs: add missing optional packages in install.md

  #2412 merged Aug 4, 2025

29 Pull requests opened by 24 people

• Use click in CLI

  #2421 opened Aug 4, 2025

• Add `on_complete` callback to AG-UI functions to get access to `AgentRunResult`

  #2429 opened Aug 5, 2025

• Add Textual TUI

  #2462 opened Aug 7, 2025

• [Example]: Multimodal RAG with LanceDB

  #2498 opened Aug 11, 2025

• Beginning of Text Editor

  #2552 opened Aug 14, 2025

• Handle function calls without text

  #2557 opened Aug 14, 2025

• Add Bedrock/Anthropic prompt caching

  #2560 opened Aug 14, 2025

• Fix StructuredDict with nested JSON schemas using $ref

  #2570 opened Aug 15, 2025

• Fixes #2513: Add support for Free-Form Function Calling and Context Free Grammar constraints over tools

  #2572 opened Aug 15, 2025

• [Draft] Add an AgentToolset

  #2589 opened Aug 17, 2025

• Add id and finish_reason to ModelResponse

  #2590 opened Aug 17, 2025

• Add LiteLLM provider for OpenAI API compatible models

  #2606 opened Aug 19, 2025

• feat: add StackOne AI integration for Pydantic AI

  #2608 opened Aug 19, 2025

• Add basic support for UploadedFile UserContent

  #2611 opened Aug 19, 2025

• Proposition MVP integration with Outlines

  #2623 opened Aug 20, 2025

• Add tool calls limit to `UsageLimits`

  #2633 opened Aug 21, 2025

• Support DBOS Agent by wrapping run as DBOS workflows and moving model requests and MCP to DBOS steps

  #2638 opened Aug 22, 2025

• Add citations

  #2657 opened Aug 25, 2025

• Add `cost` to `RunUsage`

  #2684 opened Aug 26, 2025

• Load MCP servers from file

  #2698 opened Aug 27, 2025

• Use output_type in generate_dataset

  #2716 opened Aug 28, 2025

• [Draft] Example for sequential tool calls

  #2718 opened Aug 28, 2025

• Support Python 3.14

  #2745 opened Aug 30, 2025

• add support for previous_response_id from Responses API

  #2756 opened Sep 1, 2025

• add response id for streaming response

  #2757 opened Sep 1, 2025

• Docs: update `MCPServerStreamableHTTP` usage example to use `url` argument

  #2758 opened Sep 2, 2025

• Fix bug related to Azure OpenAI model streaming response

  #2763 opened Sep 2, 2025

• Add `operation.cost` span attribute to model request spans

  #2767 opened Sep 2, 2025

• Move `mcp-run-python`

  #2776 opened Sep 2, 2025

212 Issues closed by 25 people

• Request: Review auto-generated MCP permission manifest for Python_Sandbox

  #2750 closed Sep 2, 2025

• UrlContextTool: '400 Tool use with function calling is unsupported' if `output_type` is specified

  #2761 closed Sep 2, 2025

• Groq models: Handling of invalid tool-calls

  #2759 closed Sep 2, 2025

• ThinkingPart not handling separate OpenAI thinking events

  #2149 closed Sep 2, 2025

• `NativeOutput` for groq Agents using supported models

  #2764 closed Sep 2, 2025

• Streaming token usage shows doubled input tokens vs non-streaming

  #2770 closed Sep 2, 2025

• Recommended way to view full requests to model (without a Logfire account)?

  #2769 closed Sep 2, 2025

• Support serializable object as input to agent run

  #2329 closed Sep 2, 2025

• `McpError` when `run-mcp-python` code sandbox handles concurrent requests

  #2655 closed Sep 2, 2025

• `genai-prices` is not a dev dependency

  #2760 closed Sep 2, 2025

• Change out storage type on pydantic a2a worker

  #2239 closed Sep 2, 2025

• Minor sandbox security issues in MCP-Run-Python

  #2332 closed Sep 2, 2025

• [Feature Request] Stateful Python Execution Environment for Agents in run_python MCP server (Jupyter-like Workspace)

  #2070 closed Sep 2, 2025

• Google Service Account authentication still tries to use Application Default Credentials

  #2746 closed Sep 1, 2025

• `AsyncTenacityTransport` does not release connections on retries

  #2724 closed Sep 1, 2025

• Assertion error when using Gemini Models in streaming mode ALONG with CodeExecutionTool()

  #2626 closed Sep 1, 2025

• GPT-5/o4-mini empty summary responses and AG-UI integration

  #2687 closed Sep 1, 2025

• Recursively merge model settings

  #2578 closed Sep 1, 2025

• Human-in-the-Loop behaviour

  #2747 closed Sep 1, 2025

• Drop 3.9 support and make dataclasses and functions with many arguments keywords-only

  #2720 closed Aug 31, 2025

• Use new OpenTelemetry GenAI chat span attribute conventions

  #2348 closed Aug 29, 2025

• Handle errors more gracefully in pydantic-evals

  #2612 closed Aug 29, 2025

• Parallel tool call containing both ToolReturn and deferred tool does not work

  #2500 closed Aug 29, 2025

• Human-in-the-loop approval of tool calls

  #1995 closed Aug 29, 2025

• Deprecate `OpenAIModelProfile.openai_supports_sampling_settings` in favor of `openai_unsupported_model_settings`

  #2721 closed Aug 29, 2025

• Message History is not working with redis.

  #2664 closed Aug 29, 2025

• `pydantic-graph`: persistence across `GraphRun`

  #1361 closed Aug 29, 2025

• Builtin tools for Anthropic models?

  #2722 closed Aug 29, 2025

• Questions about message mechanism

  #2654 closed Aug 28, 2025

• Providing API Key programmatically

  #2706 closed Aug 28, 2025

• Function tools don't support datetime arguments

  #2682 closed Aug 28, 2025

• Agent with MCP in google Colab

  #2697 closed Aug 28, 2025

• Incorrect PartStart classification in some OpenAI models

  #2580 closed Aug 27, 2025

• AttributeError: 'NoneType' object has no attribute 'reasoning_tokens' with vLLM +OpenAIResponsesModel

  #2598 closed Aug 26, 2025

• Could we add message history to the agent.to_cli() ?

  #2650 closed Aug 26, 2025

• Tool function with `BaseModel` argument does not work with Temporal

  #2666 closed Aug 26, 2025

• openai.BadRequestError: Invalid type for 'messages[1].content': expected one of a string or array of objects, but got an object instead.

  #2653 closed Aug 26, 2025

• Missing media_type

  #2675 closed Aug 26, 2025

• Tool not executed - a final result was already processed.

  #2679 closed Aug 26, 2025

• Control Usage Limit Exceeded Behaviour

  #2681 closed Aug 26, 2025

• Gracefully handling tool failures

  #2671 closed Aug 26, 2025

• Add Cerebras to models.

  #1826 closed Aug 26, 2025

• `AsyncTenacityTransport` throws `The retry controller did not make any attempts` when concurrent requests are retried

  #2661 closed Aug 26, 2025

• `validate_response` handler in `AsyncTenacityTransport` receives a `Response` without `request` property

  #2656 closed Aug 26, 2025

• Documented way to get agent deps from graph run?

  #2602 closed Aug 26, 2025

• Qwen3-coder usage Error caused by non numeric data in the return value

  #2632 closed Aug 26, 2025

• 🐛 MCP Tool Serialization/Deserialization Inconsistency: tool-return vs builtin-tool-return

  #2649 closed Aug 26, 2025

• Langfuse tracing customization

  #2665 closed Aug 25, 2025

• UnexpectedModelBehavior with Google Gemini thinking mode: "Received empty model response"

  #2480 closed Aug 25, 2025

• Issue with Bedrock and Claude Model: ValidationException on Blank Text Field

  #1678 closed Aug 25, 2025

• Tool calling issues with vllm

  #224 closed Aug 25, 2025

• Error in multi-agent example: pydantic_ai.exceptions.UsageLimitExceeded: Exceeded the total_tokens_limit of 300 (total_tokens=353)

  #1308 closed Aug 25, 2025

• Should we move to aioboto3?

  #1222 closed Aug 25, 2025

• (openai.APITimeoutError) OpenAI requests time out, but Gemini/Anthropic requests work fine

  #774 closed Aug 25, 2025

• Agent delegation: Agent does not retake control after delegate agent finishes when using run_stream()

  #1247 closed Aug 25, 2025

• Add iteration and verifier param to agent class for ReAct capability

  #972 closed Aug 25, 2025

• Combination of base64 file encoding and function calling returns 500 from OpenAI

  #2642 closed Aug 25, 2025

• How to make Pydantic AI output model intent as text before tool call (like Claude Desktop)?

  #2525 closed Aug 23, 2025

• MCP package conflicts with internal mcp module

  #2648 closed Aug 23, 2025

• Getting "Received empty model response" when using GPT 5 model series

  #2644 closed Aug 22, 2025

• Should we have a higher number in the pytest -n auto ? auto currently assigns to 4 workers

  #2547 closed Aug 22, 2025

• Llm tokens cost calculation ignores thoughtsTokenCount

  #2622 closed Aug 21, 2025

• Race Condition in GoogleVertexProvider Authentication Refresh Leading to AssertionError with Concurrent Requests

  #1440 closed Aug 21, 2025

• How structured `output_type` is implemented?

  #2620 closed Aug 21, 2025

• UnexpectedModelBehavior exception raised unnecessarily for streaming Gemini response

  #2630 closed Aug 21, 2025

• Add ability to load conversation history within specified date ranges in CLI

  #1981 closed Aug 20, 2025

• Invalid file type: text/csv

  #2599 closed Aug 20, 2025

• Agent fails with http 400 in streaming mode, works fine in non-streaming mode for Gemini models

  #1439 closed Aug 20, 2025

• Cannot upload images to gemini

  #1082 closed Aug 20, 2025

• Multi agent stream output without any content, run can output

  #1127 closed Aug 20, 2025

• Allow customising GCP authentication for Vertex (for using workload identity federation)

  #1520 closed Aug 20, 2025

• Anthropic tool response nested content blocks are stringified (no image response possible?)

  #1267 closed Aug 20, 2025

• Feature Request: Realtime/Live API integration

  #2113 closed Aug 20, 2025

• LLM Request Timeout

  #1438 closed Aug 20, 2025

• Intermittent 400 Errors with Tool Calling in Groq API (qwen-qwq-32b)

  #1233 closed Aug 20, 2025

• Parsing tool calls and responses handling for models in multi step tool calling

  #1163 closed Aug 20, 2025

• tavily fallback if duck duck go fails

  #1166 closed Aug 20, 2025

• how to build codeact agent

  #1393 closed Aug 20, 2025

• openai.APITimeoutError: Request timed out.

  #2619 closed Aug 20, 2025

• Anthropic requires the thinking block alongside the tool return

  #2621 closed Aug 20, 2025

• Rare Assertion error when using Gemini Models in streaming mode

  #2609 closed Aug 20, 2025

• Filter thoughts from responses for Gemini Models

  #2018 closed Aug 20, 2025

• Agent output type cannot handle dictionary

  #2603 closed Aug 20, 2025

• Get dependencies in function Tool.from_schema()

  #2122 closed Aug 20, 2025

• Why is mypy and pyright used for pydantic AI?

  #2613 closed Aug 20, 2025

• type error when trying to process `agent: Agent[None, BaseModel]` as a type on `BaseModel`

  #2471 closed Aug 20, 2025

• `Tool.from_schema` should take `takes_ctx`

  #2614 closed Aug 20, 2025

• Building graph out of agent iter

  #1467 closed Aug 19, 2025

• Best practices for managing tokens in context with tool calling

  #1495 closed Aug 19, 2025

• Node attributes vs state

  #1521 closed Aug 19, 2025

• Customizable Thinking tags

  #2334 closed Aug 18, 2025

• HTTP/2 for direct model requests

  #2577 closed Aug 18, 2025

• Using the Gemini model via Vertex ignores provider and uses GLA instead

  #1601 closed Aug 18, 2025

• Cohere usage is broken

  #2495 closed Aug 18, 2025

• Test fail with new mistralai version

  #2082 closed Aug 18, 2025

• Support tools defined on Pydantic models

  #2479 closed Aug 18, 2025

• How to set repetition penalty?

  #1460 closed Aug 18, 2025

• Add OpenAI Whisper support

  #1362 closed Aug 18, 2025

• Bedrock EU models are missing

  #2376 closed Aug 18, 2025

• `additionalProperties` is not supported by Gemini

  #2588 closed Aug 18, 2025

• [OpenAIModel] Qwen2.5 assistant output on tool call is empty

  #1446 closed Aug 18, 2025

• Support ACP protocol

  #1742 closed Aug 17, 2025

• Graph Edge Mapping / Looping?

  #1428 closed Aug 17, 2025

• RuntimeError: Event loop is closed when running official MCP Run Python example code

  #1454 closed Aug 16, 2025

• Differentiation between use of Server Tools and Client Tools

  #1459 closed Aug 16, 2025

• apply llm corection on pydantic parse error

  #2575 closed Aug 16, 2025

• Using Mistral Agents with tool and message_history leads to "Unexpected role 'user' after role 'tool'"

  #1210 closed Aug 15, 2025

• Notes on structured output strategies

  #1615 closed Aug 15, 2025

• Allow users to add more mcp servers during a chat app lifecycle

  #1300 closed Aug 15, 2025

• OpenAIModel fails with OpenRouter Gemini: missing "choices" and "created" keys

  #1746 closed Aug 15, 2025

• Running many models on Bedrock with `output_type` fails with "This model doesn't support the toolConfig.toolChoice.any field."

  #2091 closed Aug 15, 2025

• add the ability to set a list of allowed tools on an mcp server

  #1219 closed Aug 15, 2025

• Agent run regenerates a dynamic system-prompt if it is missing in 'message_history'

  #1211 closed Aug 15, 2025

• Register new tools during a run

  #1814 closed Aug 15, 2025

• Unsuported role system for openai-o1-mini

  #974 closed Aug 15, 2025

• FallbackModel type annotations does not accept string as model leading to type checking issues

  #2563 closed Aug 15, 2025

• Token Counting

  #2567 closed Aug 15, 2025

• o1 Model fails when making tool calls

  #761 closed Aug 15, 2025

• Claude Sonnet 3.5 & 4.0 fail to call tools in streaming mode (v0.2.14)

  #1904 closed Aug 15, 2025

• Integrate genai-prices

  #2352 closed Aug 15, 2025

• Support OpenAI web search tool events in streaming mode

  #1810 closed Aug 15, 2025

• Support for Dynamic Instructions

  #1749 closed Aug 15, 2025

• How to handle large numbers of tools?

  #1664 closed Aug 15, 2025

• Configuration and parameters for `all_messages()` and `new_messages()`

  #496 closed Aug 15, 2025

• Implement filtering for MCP tools

  #1432 closed Aug 15, 2025

• Assert should be used on this

  #1390 closed Aug 15, 2025

• Possible outdated document on state persistence

  #1324 closed Aug 15, 2025

• Weird tool call arguments, resulting in UnexpectedModelBehaviour / validation error

  #81 closed Aug 15, 2025

• Structured output validation keeps failing

  #1192 closed Aug 15, 2025

• Allow third-party JSON Schemas to be specified for `final_result` / output

  #1194 closed Aug 15, 2025

• run_stream docs are a little misleading

  #656 closed Aug 15, 2025

• Using multiple Agents to handle different user sessions

  #1209 closed Aug 15, 2025

• Plan and Execute is not working inside one Agent

  #985 closed Aug 15, 2025

• Appending description to the return type fields when calling models

  #851 closed Aug 15, 2025

• Feat: Add ability to differentiate tool-calling models

  #842 closed Aug 15, 2025

• [Feature] Provide a hook to customize the user prompt similar to how we do it for dynamic system prompts

  #801 closed Aug 15, 2025

• Retries Ineffective in Resolving Errors

  #739 closed Aug 15, 2025

• Need help orchestrating agents (could be a bug?)

  #1069 closed Aug 15, 2025

• Timing of model calls

  #2565 closed Aug 14, 2025

• WebSearchTool max_uses parameter ignored in Anthropic model integration

  #2555 closed Aug 14, 2025

• Missing Tokens from Beginning of any Streaming text content after 0.4.*

  #2428 closed Aug 14, 2025

• How to stream progress from tool agent calls in nested agent setups?

  #2356 closed Aug 14, 2025

• anthropic-beta and VertexAI

  #2512 closed Aug 14, 2025

• Update Agent Signature in lieu of PEP-747

  #1648 closed Aug 14, 2025

• Support for Gemini 2.5 code_execution tool

  #1776 closed Aug 14, 2025

• Add LangChain Tool support

  #1831 closed Aug 14, 2025

• TypeError: unhashable type: 'list'

  #1869 closed Aug 14, 2025

• __init__.py in the root directory is not created for some reason

  #1902 closed Aug 14, 2025

• Need custom endpoint support for Bedrock

  #2057 closed Aug 14, 2025

• Anthropic Web Search Tool

  #1683 closed Aug 14, 2025

• Behavior of Union Types as Output type

  #2509 closed Aug 14, 2025

• Support for gpt5's 'verbosity' option

  #2478 closed Aug 14, 2025

• Fireworks exception with images

  #2424 closed Aug 14, 2025

• Default `anthropic-beta` feature set by AnthropicModel incompatible with other providers

  #2543 closed Aug 14, 2025

• Remove UserWarning: Pydantic serializer warnings when using dataset.evaluate_sync without logfire installed

  #2515 closed Aug 14, 2025

• Improved ergonomics for Agent experimentation and evaluation

  #1363 closed Aug 14, 2025

• Metadata in Eval description in Dataset

  #1589 closed Aug 14, 2025

• Pydantic Eval: Evaluator does not support list of int

  #1431 closed Aug 14, 2025

• evals progress bar

  #1657 closed Aug 14, 2025

• `ModelSettings` merging not working when wrapped with `FallbackModel`

  #2539 closed Aug 13, 2025

• Access message history when UsageLimitExceeded is raised

  #1083 closed Aug 13, 2025

• Feature Request: Add Evaluation Support Mention in Confident AI Integration Docs

  #2461 closed Aug 13, 2025

• Native temporal support

  #1975 closed Aug 12, 2025

• Adding an option in`StreamedRunResult.stream_structured()` to include the tool call results

  #2492 closed Aug 12, 2025

• best practice that allows an agent to output a natural dialogue text and then make tool calls?

  #2411 closed Aug 12, 2025

• KeyError: 'token_count' when using Gemini with VideoUrl

  #2489 closed Aug 12, 2025

• Feature request: Add current messages to UsageLimitExceeded error

  #931 closed Aug 12, 2025

• Questions about end_strategy, is it being used? How does it work?

  #2510 closed Aug 12, 2025

• How to stream tool usage and final answers in MCP calls with pydantic-ai?

  #2497 closed Aug 12, 2025

• Using Function Model with streaming responses

  #2505 closed Aug 12, 2025

• Latest openai package fails with TypeError: Cannot instantiate typing.Union

  #2476 closed Aug 12, 2025

• Working MCP client tools with openai azure hosted models

  #2517 closed Aug 12, 2025

• Clarification on .iter() vs .iter_from_persistence()

  #2460 closed Aug 12, 2025

• expose `_agent_graph.AgentNode` via a public api

  #2474 closed Aug 12, 2025

• ModuleNotFoundError after release logfire 4.3.0

  #2516 closed Aug 12, 2025

• Logfire doesn't render numeric tool return value

  #2363 closed Aug 12, 2025

• Pydantic AI doesn't work with openai>=1.99.2 due to changes in the tool types mappings

  #2504 closed Aug 12, 2025

• The `Agent.system_prompt` is discarded when `message_history` contain `ModelResponse`

  #1032 closed Aug 11, 2025

• Save `history_processors`'s result for next round model request

  #2095 closed Aug 11, 2025

• GPT-5 fail gracefully if temperature parameter is provided

  #2482 closed Aug 11, 2025

• Error setting gpt 5 'minimal' thinkingeffort

  #2494 closed Aug 11, 2025

• In evals, how to detect whether a tool was called?

  #2212 closed Aug 9, 2025

• MCP argument injection/session based header

  #2484 closed Aug 8, 2025

• Dynamic toolsets based on run context

  #2331 closed Aug 8, 2025

• Feature Request - Toolset for working with Excel files

  #2473 closed Aug 8, 2025

• Add Agent `event_stream_handler` function as alternative to `Agent.iter`

  #1977 closed Aug 8, 2025

• Make FunctionToolsetTool public

  #2464 closed Aug 7, 2025

• mcp-run-python jsr release not up to date

  #2442 closed Aug 7, 2025

• Support built-in tools

  #840 closed Aug 7, 2025

• object of type 'NotGiven' has no len() after ugrade from 0.4.11 to 0.5.0

  #2434 closed Aug 7, 2025

• help on playwright mcp + pydantic-ai task

  #2448 closed Aug 7, 2025

• JSON schema concerns: annotations and descriptions sometimes ignored

  #2405 closed Aug 6, 2025

• Adding CodeAgents; Agents that take actions in Code

  #2380 closed Aug 6, 2025

• Response caching

  #1093 closed Aug 6, 2025

• Support Amazon Bedrock API Keys

  #2159 closed Aug 6, 2025

• unhandled errors in a TaskGroup (1 sub-exception) when use pyinstaller with --noconsole

  #2432 closed Aug 6, 2025

• Bedrock + Anthropic extended thinking: error because thinking parts are not sent back

  #2453 closed Aug 6, 2025

• The output_type does not support setting object lists, but it is possible to use the SDK of OpenAI

  #2437 closed Aug 6, 2025

• Anthropic extended thinking: invalid request error when using tools

  #2040 closed Aug 6, 2025

• Would you guys be open to add caching with joblib memory?

  #2417 closed Aug 6, 2025

• Remove deprecated arguments and methods from before April ahead of v1

  #2350 closed Aug 6, 2025

• Deprecate GeminiModel in favor of GoogleModel

  #2351 closed Aug 6, 2025

• mcp-run-js exists ?

  #2259 closed Aug 5, 2025

• Unit Testing a TestModel Agent with MCP Toolset Access?

  #2296 closed Aug 5, 2025

• Docs: Tools become FunctionToolset

  #2385 closed Aug 4, 2025

• DocumentUrl in message_history is missing "_media_type"

  #2394 closed Aug 4, 2025

• Can't save a EvaluationReport to json and load it again

  #1953 closed Aug 4, 2025

• Can you make it easier to add the OpenAI built in code interpreter for the OpenAI Responses API?

  #2392 closed Aug 4, 2025

• Question: Why do Agent responses include ToolCallPart and ToolReturnPart in ModelRequest output?

  #2399 closed Aug 4, 2025

• How to use Agent.to_ag_ui in stdio mode

  #2400 closed Aug 4, 2025

• test_download_item_no_content_type test fails on macOS

  #2403 closed Aug 4, 2025

• type error while assigning openrouter provider as model provider

  #2407 closed Aug 4, 2025

50 Issues opened by 43 people

• to_cli_sync: when agent is provided with `output_type` of a function, the function is called multiple times with partial input parameters

  #2773 opened Sep 2, 2025

• AG UI: Validation errors when using reasoning models with Ollama (gpt-oss and qwen3)

  #2766 opened Sep 2, 2025

• Can POML be used in PydanticAI?

  #2751 opened Sep 1, 2025

• Support MCP tool call response `structuredContent`

  #2742 opened Aug 30, 2025

• Z.ai models

  #2723 opened Aug 29, 2025

• Magistral - handling of mistralai.models.thinkchunk.ThinkChunk

  #2713 opened Aug 28, 2025

• Use a better OTel attribute structure for evals case data

  #2708 opened Aug 28, 2025

• Enable: output_type + stream_text

  #2707 opened Aug 28, 2025

• Send ThinkingPart back for OpenAIChatModel models which require reasoning content

  #2701 opened Aug 27, 2025

• Potential MCP issue (asyncio.exceptions.CancelledError)

  #2700 opened Aug 27, 2025

• Passing in tools into the Evals LLM judge

  #2667 opened Aug 25, 2025

• OpenAI Responses: `previous_response_id` support.

  #2663 opened Aug 25, 2025

• Loss of Final Tokens After Tool Call When Using kimi-k2 Model with Streaming Output

  #2660 opened Aug 25, 2025

• Missing Initial Tokens After Tool Call in Streaming Output (But Present in Final Output) with Gemini 2.5 Flash Model Using agent.iter()

  #2659 opened Aug 25, 2025

• deepseek support

  #2651 opened Aug 24, 2025

• Add "modelretry" end strategy

  #2646 opened Aug 22, 2025

• Getting "Received empty model response" when using GPT 5 model series

  #2645 opened Aug 22, 2025

• Failed to yield ToolReturnPart streaming with structured output type via tool call

  #2640 opened Aug 22, 2025

• Expose Span Customization / Event Iteration for handle_ag_ui_request and other AG-UI functions

  #2639 opened Aug 22, 2025

• Add identifier field to FileUrl and subclasses

  #2635 opened Aug 21, 2025

• DBOS Support

  #2629 opened Aug 20, 2025

• Run Tools Sequentially

  #2628 opened Aug 20, 2025

• How to access metadata added in `MCPServer.process_tool_call` from MCP server tool

  #2616 opened Aug 20, 2025

• MCPServerStdio initialization exception not propagated

  #2607 opened Aug 19, 2025

• How to define a fallback function when UsageLimits are exceeded?

  #2605 opened Aug 19, 2025

• on sporadic `ModelHTTPError` when using web_search builtin_tools with Anthropic

  #2600 opened Aug 18, 2025

• Streaming mode doesn't capture LLM's thought recovery from chunk.choices[0].delta.model_extra

  #2595 opened Aug 18, 2025

• Send back reasoning tokens to LLM within user turn

  #2594 opened Aug 18, 2025

• Enforce usage limits by exact tool call count (not just request/turn count)

  #2593 opened Aug 18, 2025

• Allow FunctionToolset() to accept default for require_parameter_descriptions

  #2587 opened Aug 16, 2025

• Function Toolset should catch exceptions and raise model retry

  #2586 opened Aug 16, 2025

• Support referencing file_ids previously uploaded to model providers.

  #2574 opened Aug 15, 2025

• How to track LLM timing

  #2566 opened Aug 14, 2025

• Anthropic max_tokens is 4096 by default

  #2553 opened Aug 14, 2025

• Processing Message History Event

  #2541 opened Aug 13, 2025

• Streaming and Tools does not work with Anthropic

  #2521 opened Aug 12, 2025

• Empty dynamic system prompt should not add a new message

  #2519 opened Aug 12, 2025

• Add support for Free-Form Function Calling and Context Free Grammar constraints over string output

  #2513 opened Aug 12, 2025

• Can not send BinaryContent to "gpt-4.1-mini" on AzureOpenAI

  #2511 opened Aug 12, 2025

• MCP client using httpx.client is closed after one request

  #2502 opened Aug 11, 2025

• ChatCompletionService giving out error when using hugging face router with providers

  #2499 opened Aug 11, 2025

• Fix naming of MCP Client & Server

  #2496 opened Aug 10, 2025

• Error when using GPT-5 with `output_type` and built-in tools because `tool_choice=required` is not supported

  #2488 opened Aug 8, 2025

• Let OTel events from Logfire be deserialized into `ModelMessage`s

  #2472 opened Aug 7, 2025

• StructuredDict passed into run cannot handle nested JsonSchema

  #2466 opened Aug 7, 2025

• Cache model responses for identical requests

  #2452 opened Aug 6, 2025

• OpenAI/gpt-oss-120b - Expected 2 output messages (reasoning and final), but got 3.

  #2449 opened Aug 6, 2025

• Support gpt-oss directly via Harmony format

  #2439 opened Aug 6, 2025

• Anthropic thinking returns error: "Thinking may not be enabled when tool_choice forces tool use."

  #2425 opened Aug 5, 2025

• Instructions being inserted before System Prompt

  #2423 opened Aug 4, 2025

72 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• add XML attributes when formatting Pydantic models in prompts

  #2313 commented on Aug 29, 2025 • 22 new comments

• Add Streaming Support to A2A

  #2362 commented on Aug 14, 2025 • 0 new comments

• Vector search and embeddings API

  #58 commented on Aug 26, 2025 • 0 new comments

• AG UI - access to event bus to pass custom events during tool run

  #2382 commented on Aug 25, 2025 • 0 new comments

• should cancel the response when user stop consuming

  #1524 commented on Aug 25, 2025 • 0 new comments

• Handoffs / sub-agent delegation

  #1978 commented on Aug 25, 2025 • 0 new comments

• Anthropic prompt caching (inc. Anthropic on Bedrock)

  #1041 commented on Aug 25, 2025 • 0 new comments

• Unable to send only BinaryContent to Gemini using Vertex

  #1409 commented on Aug 25, 2025 • 0 new comments

• Using BinaryContent with CSV media type throws an error

  #1161 commented on Aug 25, 2025 • 0 new comments

• Streaming final output while iterating over agent graph

  #1449 commented on Aug 21, 2025 • 0 new comments

• Define complex eval case through code in a pytest-class fashion

  #2287 commented on Aug 20, 2025 • 0 new comments

• Support both tool call and message text handling in response

  #675 commented on Aug 20, 2025 • 0 new comments

• Get Model response even when result_type is not expected

  #708 commented on Aug 20, 2025 • 0 new comments

• MCP improvement: `mcp-run-python` should be be able to call back to the client

  #1271 commented on Aug 20, 2025 • 0 new comments

• Calculating custom aggregate metrics for evals

  #1413 commented on Aug 20, 2025 • 0 new comments

• gpt-4o: `parallel_tool_calls` calls too many tools, fails with `Expected an array with maximum length 128`

  #1105 commented on Aug 20, 2025 • 0 new comments

• toolcalling with vllm is broken

  #1414 commented on Aug 19, 2025 • 0 new comments

• Access Starlette request in tool calls

  #2264 commented on Aug 19, 2025 • 0 new comments

• `validate_structured_result` returns empty while usage shows there's an output (also it doesn't go to except)

  #896 commented on Aug 18, 2025 • 0 new comments

• Fix MCP client documentation inconsistencies

  #2357 commented on Aug 14, 2025 • 0 new comments

• Capture thought signatures from Google models

  #2305 commented on Aug 14, 2025 • 0 new comments

• Added MCP elicitation support for tool injection in mcp-run-python

  #2258 commented on Sep 2, 2025 • 0 new comments

• clai: Add ability to continue last conversation

  #2257 commented on Aug 14, 2025 • 0 new comments

• Add provider for Anthropic's Vertexai Client

  #1392 commented on Aug 21, 2025 • 0 new comments

• Endless Tool Loop (Vertex Gemini)

  #765 commented on Sep 2, 2025 • 0 new comments

• Add `id` and `finish_reason` to `ModelResponse`

  #886 commented on Sep 1, 2025 • 0 new comments

• Gemini causes 'Event loop is closed' when running inside an async context

  #748 commented on Sep 1, 2025 • 0 new comments

• Azure OpenAI API Streaming Response Causes AttributeError in pydantic-ai

  #797 commented on Sep 1, 2025 • 0 new comments

• Pydantic AI V1

  #1974 commented on Aug 31, 2025 • 0 new comments

• Ability to Persist Messages in External Stores

  #530 commented on Aug 30, 2025 • 0 new comments

• How do I handle pydantic_ai.exceptions.UnexpectedModelBehavior

  #1380 commented on Aug 29, 2025 • 0 new comments

• Gemini unable to stream structured output

  #1237 commented on Aug 27, 2025 • 0 new comments

• Add a llama-cpp Model?

  #1801 commented on Aug 27, 2025 • 0 new comments

• How to use streaming in sync ?

  #1502 commented on Aug 26, 2025 • 0 new comments

• Question on Batching (for Anthropic Models)

  #1144 commented on Aug 26, 2025 • 0 new comments

• More capability for dynamic system prompt (depending on tools and context providers)

  #700 commented on Aug 26, 2025 • 0 new comments

• Messages passed to history_processor functions include current run messages

  #2050 commented on Aug 14, 2025 • 0 new comments

• Gemini 2.5 Pro Streamed response has no content field (just streamed thinking parts)

  #2097 commented on Aug 14, 2025 • 0 new comments

• Getting Intermediate responses from MCP Tool Calls

  #1748 commented on Aug 14, 2025 • 0 new comments

• Thought Signatures With Google Gemini Models

  #2293 commented on Aug 14, 2025 • 0 new comments

• synergy between graph and evals

  #1906 commented on Aug 14, 2025 • 0 new comments

• Create Agents with Oracle Cloud Infrastructure (OCI) Hosted LLMs

  #1385 commented on Aug 14, 2025 • 0 new comments

• Gemini Agent run from cached context

  #664 commented on Aug 13, 2025 • 0 new comments

• CI is not testing evals without logfire

  #1401 commented on Aug 12, 2025 • 0 new comments

• Force Agent tool use on first request

  #1820 commented on Aug 12, 2025 • 0 new comments

• Reinforcement Learning for Pydantic AI Agents

  #2202 commented on Aug 8, 2025 • 0 new comments

• Feature Request: Synchronous Calls

  #934 commented on Aug 7, 2025 • 0 new comments

• TimeoutError in MCPServerStreamableHTTP

  #2401 commented on Aug 6, 2025 • 0 new comments

• No way to stop the agent after using a tool and return the tool response

  #508 commented on Aug 6, 2025 • 0 new comments

• How can one simply return the response of the tool, instead of routing the response to a final result handler?

  #127 commented on Aug 6, 2025 • 0 new comments

• Support Python 3.14

  #2402 commented on Aug 5, 2025 • 0 new comments

• Callback/Middleware for responses on `run_ag_ui`

  #2398 commented on Aug 5, 2025 • 0 new comments

• Contributing FastMCP Toolset

  #2406 commented on Aug 4, 2025 • 0 new comments

• agent.run_stream doesn't run with tools

  #2308 commented on Aug 4, 2025 • 0 new comments

• MCP embedded resource metadata is not passed to model

  #2288 commented on Aug 18, 2025 • 0 new comments

• How can I export file(s) from Pyodide to local when using MCP Run Python?

  #1455 commented on Aug 16, 2025 • 0 new comments

• Support "name" for naming chat messages with the same role.

  #1863 commented on Aug 16, 2025 • 0 new comments

• Token Counter

  #1106 commented on Aug 16, 2025 • 0 new comments

• Cost accounting

  #1443 commented on Aug 16, 2025 • 0 new comments

• Duplicated results when streaming with tool calls

  #1391 commented on Aug 15, 2025 • 0 new comments

• OpenAI Built In Code Interpreter not returning image files from OpenAI Response

  #2391 commented on Aug 15, 2025 • 0 new comments

• AWS Bedrock Inference Profile Support

  #1493 commented on Aug 15, 2025 • 0 new comments

• Missing MCP feature: MCP Tool Annotations are not returned in ToolDefinition from MCP Servers

  #1526 commented on Aug 15, 2025 • 0 new comments

• Nesting capture_run_messages

  #1568 commented on Aug 15, 2025 • 0 new comments

• Upgrade Pyodide after next release

  #1992 commented on Aug 15, 2025 • 0 new comments

• Feature Request: provide an easy way to include your (versioned) API docs in LLM contexts

  #944 commented on Aug 15, 2025 • 0 new comments

• logfire within examples

  #1278 commented on Aug 15, 2025 • 0 new comments

• Feature request: Provide a better way to override the result of `Agent.run` in testing

  #872 commented on Aug 15, 2025 • 0 new comments

• `pydantic-ai-graph` feature comparison / parity chart?

  #804 commented on Aug 15, 2025 • 0 new comments

• How to extract more tokens than the model's max for a single call.

  #718 commented on Aug 15, 2025 • 0 new comments

• Pydantic Evals report can't be displayed if logfire is not installed

  #2209 commented on Aug 15, 2025 • 0 new comments

• Adding LiteLLM as model wrap just like how google-adk does it.

  #1496 commented on Aug 15, 2025 • 0 new comments