- Notifications
You must be signed in to change notification settings - Fork1k
Insights: pydantic/pydantic-ai
Overview
Could not load contribution data
Please try again later
2 Releases published by2 people
- v0.4.3 v0.4.3 (2025-07-16)
published
Jul 16, 2025 - v0.4.4 v0.4.4 (2025-07-18)
published
Jul 18, 2025
29 Pull requests merged by18 people
- chore: Fix inconsistent docs example output and fix variable name
#2248 merged
Jul 18, 2025 - Support AG-UI protocol for frontend-agent communication
#2223 merged
Jul 18, 2025 - Add grok-4 and groq kimi-k2 models
#2235 merged
Jul 18, 2025 - add identifier field to BinaryContent class
#2231 merged
Jul 18, 2025 - Toolsets public interface and docs tweaks
#2241 merged
Jul 17, 2025 - chore(mistral): disable model_fields deprecation warning
#2224 merged
Jul 17, 2025 - feat: update OpenAI models - add o1-pro, o3-pro, o3-deep-research, computer-use
#2234 merged
Jul 17, 2025 - Nicer errors under the capture_run_messages context
#2219 merged
Jul 17, 2025 - ⚡️ Speed up method
Usage.opentelemetry_attributesby 85%#2198 merged
Jul 17, 2025 - ⚡️ Speed up method
AgentRunResult._set_output_tool_returnby 18798%#2196 merged
Jul 17, 2025 - update logos, favicon and brand names
#2193 merged
Jul 17, 2025 - chore: add experimetal GitHub agent
#2232 merged
Jul 17, 2025 - Fix VertexAI Empty Model Parts Error
#2203 merged
Jul 17, 2025 - change clai default model to gpt-4.1
#2227 merged
Jul 17, 2025 - [Docs] List Hugging Face Inference Provider
#2229 merged
Jul 17, 2025 - Add Douwe to maintainers
#2221 merged
Jul 16, 2025 - Remove old Google models
#2220 merged
Jul 16, 2025 - [openai] Remove incorrect tool call id from tool call delta
#2210 merged
Jul 16, 2025 - Toolsets
#2024 merged
Jul 16, 2025 - Add Hugging Face as a provider
#1911 merged
Jul 16, 2025 - feat: Add output function tracing
#2191 merged
Jul 16, 2025 - fix: a2a docs dependency
#2216 merged
Jul 16, 2025 - Bugfix: avoid race condition when refreshing google token
#2100 merged
Jul 15, 2025 - Add base64 encoding to
tool_return_ta#2186 merged
Jul 15, 2025 duckduckgo-searchis renamed toddgs#2172 merged
Jul 15, 2025 - Mention agents instead of models in MCP servers documentation
#2160 merged
Jul 15, 2025 - docs: Add Agenta integration documentation
#2192 merged
Jul 15, 2025 - Include content.1571
#2180 merged
Jul 15, 2025 - Remove unneeded
pragma: lax no cover#2177 merged
Jul 14, 2025
15 Pull requests opened by11 people
- Install Codeflash Github action
#2195 opened
Jul 14, 2025 - ⚡️ Speed up function `_map_usage` by 172%
#2197 opened
Jul 14, 2025 - Fix include_content not working as expected
#2206 opened
Jul 15, 2025 - anthropic: drop new lines on empty system prompt
#2208 opened
Jul 15, 2025 - feat: add MoonshotAI provider with Kimi-K2 model support
#2211 opened
Jul 15, 2025 - chore: dedupe function call try block in _output.py
#2217 opened
Jul 16, 2025 - WIP: Temporal Agent
#2225 opened
Jul 17, 2025 - validate OpenAI responses
#2226 opened
Jul 17, 2025 - change `format_as_xml` defaults
#2228 opened
Jul 17, 2025 - feat: add support for streamable http in the mcp-run-python tool
#2230 opened
Jul 17, 2025 - Adding thinkingpart to otel_events in ModelResponse
#2237 opened
Jul 17, 2025 - pass sse_read_timeout to MCP ClientSession read_timeout_seconds
#2240 opened
Jul 17, 2025 - Handle `None` `created` timestamp coming from OpenRouter API
#2247 opened
Jul 18, 2025 - Add name argument to FunctionToolset, DeferredToolset and MCPServer
#2250 opened
Jul 18, 2025 - Handle tool errors better in tool registry
#2252 opened
Jul 18, 2025
22 Issues closed by9 people
- AG-UI Protocol Integration with Pydantic AI
#1864 closed
Jul 18, 2025 - Improving Tools: User vs LLM representations in the UI
#2245 closed
Jul 18, 2025 - Improving Tools: Asking for Permissions
#2246 closed
Jul 18, 2025 - BinaryContent returned by a tool is replaced with e4fcfe by agent
#2243 closed
Jul 18, 2025 - Official Support For XAI models (namely Grok 4)
#2187 closed
Jul 18, 2025 - Add optional src or metadata field to BinaryContent
#2139 closed
Jul 18, 2025 - Feature request: support for Comet Opik LLM trace evaluator
#2204 closed
Jul 17, 2025 - Improve error trace for user errors that happen inside capture_run_messages
#2218 closed
Jul 17, 2025 - GoogleModel generates empty text parts that Vertex AI rejects with 400 error
#2032 closed
Jul 17, 2025 - Remove deprecated preview models from LatestGeminiModelNames
#2016 closed
Jul 16, 2025 - Make output_type be conditionally available
#2042 closed
Jul 16, 2025 - Client tool calling support
#1817 closed
Jul 16, 2025 - `Toolset`, take 3
#1973 closed
Jul 16, 2025 - Failed to render details panel
#1567 closed
Jul 16, 2025 - Add HuggingFace to Models.
#1085 closed
Jul 16, 2025 - Gemini 2.5 Pro Thinking Budget Not Working (possible google bug)
#1967 closed
Jul 16, 2025 - Show output functions span in logfire
#2108 closed
Jul 16, 2025 - Pydantic AI duckduckgo search tool out of date
#2182 closed
Jul 16, 2025 - Local backends?
#2200 closed
Jul 16, 2025 - Add option to filter reasoning content from message history
#1157 closed
Jul 15, 2025 - v0.3.3 release breaks logfire integration for BinaryContent
#2183 closed
Jul 15, 2025 - Support arbitrary attributes in `ModelSettings` for use by OpenAI compatible providers
#1101 closed
Jul 12, 2025
18 Issues opened by17 people
- Handle tool errors better in tool registry
#2251 opened
Jul 18, 2025 - What are our options for installing the most recent dev version?
#2249 opened
Jul 18, 2025 - `<think>` tags in text parts are not parsed into thinking parts when streaming
#2244 opened
Jul 18, 2025 - Streaming Structured Output with Gemini Model over vertex-ai
#2242 opened
Jul 18, 2025 - Change out storage type on pydantic a2a worker
#2239 opened
Jul 17, 2025 - [TEST] This is a title
#2233 opened
Jul 17, 2025 - Add read_timeout_seconds to MCPServerStreamableHTTP
#2222 opened
Jul 16, 2025 - Using native clients for bedrock models
#2215 opened
Jul 16, 2025 - A2A - How to return a input-required task to user and pause execution
#2213 opened
Jul 16, 2025 - In evals, how to detect whether a tool was called?
#2212 opened
Jul 15, 2025 - Pydantic Evals report can't be displayed if logfire is not installed
#2209 opened
Jul 15, 2025 - MCP with A2A
#2207 opened
Jul 15, 2025 - Reinforcement Learning for Pydantic AI Agents
#2202 opened
Jul 14, 2025 - Multiples MCPs: AnyIO cancel scope stack corruption
#2201 opened
Jul 14, 2025 - Using TestModel to run specific output tool.
#2199 opened
Jul 14, 2025 - Support web search citation output (OpenAI Responses API)
#2194 opened
Jul 14, 2025 - Expose tool calling primitives
#2189 opened
Jul 12, 2025
39 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 `builtin_tools` to `Agent`
#2102 commented on
Jul 18, 2025 • 12 new comments - Update tests to be compatible with new OpenAI, MistralAI and MCP versions
#2094 commented on
Jul 17, 2025 • 12 new comments - Fix LLMJudge input handling to preserve BinaryContent as separate message part instead of stringifying
#2173 commented on
Jul 17, 2025 • 7 new comments - Dockerise the MCP Run Python server (#1837)
#2090 commented on
Jul 15, 2025 • 2 new comments - Add a llama-cpp Model?
#1801 commented on
Jul 13, 2025 • 0 new comments - Question on Batching (for Anthropic Models)
#1144 commented on
Jul 17, 2025 • 0 new comments - support batch processing
#1771 commented on
Jul 17, 2025 • 0 new comments - Support `o3-deep-research` & `o4-mini-deep-research`
#2086 commented on
Jul 18, 2025 • 0 new comments - tool calling with MCP does not work for gemma3 with pydantic-ai
#2170 commented on
Jul 18, 2025 • 0 new comments - Save `history_processors`'s result for next round model request
#2095 commented on
Jul 18, 2025 • 0 new comments - Human-in-the-loop approval of tool calls
#1995 commented on
Jul 18, 2025 • 0 new comments - Django persistence integration
#2185 commented on
Jul 18, 2025 • 0 new comments - Support passing files uploaded to Gemini Files API
#2005 commented on
Jul 18, 2025 • 0 new comments - Add `tool_choice` to `ModelSettings`
#825 commented on
Jul 15, 2025 • 0 new comments - Add provider for Anthropic's Vertexai Client
#1392 commented on
Jul 15, 2025 • 0 new comments - Improve docs index and README
#2058 commented on
Jul 15, 2025 • 0 new comments - Adding CountToken to Gemini
#2137 commented on
Jul 16, 2025 • 0 new comments - CLI refactor
#2138 commented on
Jul 16, 2025 • 0 new comments - ⚡️ Speed up function `_estimate_string_tokens` by 221%
#2156 commented on
Jul 16, 2025 • 0 new comments - Fix multi reasoning_summary for openai
#2162 commented on
Jul 17, 2025 • 0 new comments - Bedrock Prompt Caching
#1845 commented on
Jul 14, 2025 • 0 new comments - Support for ["cache-control"] = {"type": "ephemeral"} in Anthropic
#1041 commented on
Jul 15, 2025 • 0 new comments - Handoffs / sub-agent delegation
#1978 commented on
Jul 15, 2025 • 0 new comments - Dynamic System prompts not being included on fresh runs with message_history
#1646 commented on
Jul 15, 2025 • 0 new comments - Add a setting to remove prompts and completions from tracing
#1571 commented on
Jul 15, 2025 • 0 new comments - Structured Output Streaming Returning as Single Chunk(Bedrock Anthropic)
#1889 commented on
Jul 15, 2025 • 0 new comments - AWS Bedrock Inference Profile Support
#1493 commented on
Jul 16, 2025 • 0 new comments - Support Anthropic tool result citations
#2128 commented on
Jul 16, 2025 • 0 new comments - ModelResponsePart should support multimodal data
#2140 commented on
Jul 16, 2025 • 0 new comments - Gemini causes 'Event loop is closed' when running inside an async context
#748 commented on
Jul 16, 2025 • 0 new comments - Gemini unable to stream structured output
#1237 commented on
Jul 16, 2025 • 0 new comments - Impossible to stream text responses ahead of final result when requiring structured output
#2158 commented on
Jul 16, 2025 • 0 new comments - Can you add support for ssl certificate files in MCPServerHTTP?
#1909 commented on
Jul 16, 2025 • 0 new comments - Dockerise the MCP Run Python server
#1837 commented on
Jul 16, 2025 • 0 new comments - [BUG] Error responses are not handled correctly for google openai/openrouter
#527 commented on
Jul 17, 2025 • 0 new comments - Adding LiteLLM as model wrap just like how google-adk does it.
#1496 commented on
Jul 17, 2025 • 0 new comments - MCP improvement: CLI support
#1270 commented on
Jul 17, 2025 • 0 new comments - LLM thinking blocks displayed in Logfire message view
#2131 commented on
Jul 17, 2025 • 0 new comments - Access ModelRequest-ModelResponse timestamps and duration as fields
#1996 commented on
Jul 17, 2025 • 0 new comments