v0.3.0 (2025-06-18)

Breaking Changes

See #1142 — Adds support for thinking parts.

We now convert the thinking blocks ("<think>..."</think>") in provider specific text parts to
PydanticAI ThinkingParts. Also, as part of this release, we made the choice to not send back the
ThinkingParts to the provider - the idea is to save costs on behalf of the user. In the future, we
intend to add a setting to customize this behavior.

---

What's Changed

• Support Thinking part by @Kludex in #1142

Full Changelog: v0.2.20...v0.3.0

v0.2.20 (2025-06-18)

What's Changed

• Handle McpError from MCP tool calls by @ppcantidio in #1999
• Add process_tool_call hook to MCP servers to modify tool args, metadata, and return value by @stevenh in #2000
• Respect ModelSettings.timeout in GoogleModel by @DouweM in #2006
• feat: add RunContext support to history processors by @Wh1isper in #2002

New Contributors

• @ppcantidio made their first contribution in #1999
• @stevenh made their first contribution in #2000

Full Changelog: v0.2.19...v0.2.20

v0.2.19 (2025-06-16)

What's Changed

• Proper check if callable is async by @Kludex in #1972
• More flexible method infer_provider by @hovi in #1945
• Ignore dynamic instructions returning an empty string by @giacbrd in #1961
• uprev Pyodide to 0.27.6 by @samuelcolvin in #1944
• refactor: updated tools doc with function naming: roll_die → roll_dice by @yamanahlawat in #1986
• Set Anthropic max_tokens to 4096 by default by @Kludex in #1994
• feat: add history_processors parameter to Agent for message processing by @Kludex in #1970
• Yield events for unknown tool calls by @proever in #1960
• Always set a parameters schema on a Gemini function declaration, even when it's an empty object by @DouweM in #1998

New Contributors

• @hovi made their first contribution in #1945
• @giacbrd made their first contribution in #1961
• @yamanahlawat made their first contribution in #1986
• @proever made their first contribution in #1960

Full Changelog: v0.2.18...v0.2.19

v0.2.18 (2025-06-13)

What's Changed

• Reuse last request from message history if no user prompt was provided by @DouweM in #1955
• Prevent Anthropic API errors from empty message content by @mike-luabase in #1934
• Add MCP Streamable HTTP implementation by @Kludex in #1965
• feat(openai): expose Responses API id as vendor_id by @sarunas-zebra in #1949
• Use GoogleModel instead of GeminiModel on inference by @Kludex in #1881

New Contributors

• @sarunas-zebra made their first contribution in #1949

Full Changelog: v0.2.17...v0.2.18

v0.2.17 (2025-06-12)

What's Changed

• Add token usage metrics to InstrumentedModel by @alexmojaki in #1898
• Add service_tier to OpenAIModelSettings by @empezarcero in #1923
• Allow users to supply httpx.AsyncClient in MCPServerHTTP by @mpfaffenberger in #1925
• Don't send sampling settings like temperature and top_p to OpenAI reasoning models by @DouweM in #1956
• Support field fileData (direct file URL) for GeminiModel and GoogleModel by @vricciardulli in #1136

New Contributors

• @Deriverx2 made their first contribution in #1946
• @amihalik made their first contribution in #1912
• @empezarcero made their first contribution in #1923
• @mpfaffenberger made their first contribution in #1925

Full Changelog: v0.2.16...v0.2.17

v0.2.16 (2025-06-08)

What's Changed

• Stop sharing tool retry count across all runs of the same agent by @dmontagu in #1918
• Add support for HerokuProvider by @Kludex in #1933
• Add stop_sequences to Google models by @Kludex in #1935
• Handle model response timestamps in milliseconds rather than seconds by @DouweM in #1937
• Add convenience method to use LangChain community tools by @matthewfranglen in #1832
• Infer the right type on output when callables are provided by @Kludex in #1931
• Revert pyodide to a version that doesn't print to stdout when installing packages by @samuelcolvin in #1930

New Contributors

• @matthewfranglen made their first contribution in #1832

Full Changelog: v0.2.15...v0.2.16

v0.2.15 (2025-06-05)

What's Changed

• Allow empty messages when using GoogleModel by @Kludex in #1922

Full Changelog: v0.2.14...v0.2.15

v0.2.14 (2025-06-03)

What's Changed

• Temporarily revert "Add support for MCP's Streamable HTTP transport (#1716)" by @DouweM in #1903

Full Changelog: v0.2.13...v0.2.14

v0.2.13 (2025-06-03)

What's Changed

• Make parallel_evaluation_example more robust by @DouweM in #1861
• Ensure tool call parts with custom argument model validation errors are serializable by @DouweM in #1862
• Add option to pass expected output to LLMJudge by @hinnefe2 in #1853
• Fix unexpected part error when Google model returns empty text delta by @Blue9 in #1875
• Generate ToolCallPart.tool_call_id when OpenAI-compatible API returned an empty string by @DouweM in #1892
• docs: Fix import for BaseModel in graph example code by @briann in #1894
• Add new _GeminiThoughtPart to adhere to Gemini response by @dnouri in #1897

New Contributors

• @hinnefe2 made their first contribution in #1853
• @Blue9 made their first contribution in #1875
• @briann made their first contribution in #1894
• @dnouri made their first contribution in #1897

Full Changelog: v0.2.12...v0.2.13

v0.2.12 (2025-05-29)

What's Changed

• Add vendor_id and vendor_details.finish_reason to Gemini/Google model responses by @davide-andreoli in #1800
• Fix units of sse_read_timeout timedelta by @alexmojaki in #1843
• Support functions as output_type, as well as lists of functions and other types by @DouweM in #1785
• Enhance Gemini usage tracking to collect comprehensive token data by @kiqaps in #1752
• docs: add scope when loading credentials from file in google by @Kludex in #1850
• Add support for Claude 4 Sonnet and Opus models in Bedrock by @samtin0x in #1838
• Add ModelProfile to let model-specific behaviors be configured independent of the model class by @DouweM in #1835
• Add new provider classes for Together AI, Fireworks AI, and Grok with automatic model profile selection by @DouweM in #1842
• Fix weather agent example with UI by @vp777 in #1851

New Contributors

• @davide-andreoli made their first contribution in #1800
• @samtin0x made their first contribution in #1838
• @vp777 made their first contribution in #1851

Full Changelog: v0.2.11...v0.2.12