Skip to content

pydantic-ai

v1.101.0 Feature

This release adds 3 notable features for engineering teams evaluating rollout.

Published 12d AI Agents & Assistants
✓ No known CVEs patched
Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

agent-framework genai llm pydantic python

ReleasePort's take

Light signal
editorial:auto 12d

Release v1.101.0 adds several new capabilities (pending message queue, MCP background tasks, model‑agnostic XSearch) and resolves multiple tool‑result handling bugs across Claude, Anthropic, and BaseModel integrations.

Why it matters: Apply the patch to fix silent loss of thinking results for Claude Sonnet 4.6/Opus 4.6 on Bedrock and other tool argument regressions; no version thresholds or deadlines are specified.

Summary

AI summary

Updates 🚀 Features, 🐛 Bug Fixes, and 📦 Dependencies across a mixed release.

Changes in this release

Feature Medium

Add pending message queue using ctx.enqueue and agent_run.enqueue by @DouweM.

Add pending message queue using ctx.enqueue and agent_run.enqueue by @DouweM.

Source: llm_adapter@2026-05-22

Confidence: high

 —
Feature Medium

Add support for MCP background tasks (SEP-1686) by @Alex-Resch.

Add support for MCP background tasks (SEP-1686) by @Alex-Resch.

Source: llm_adapter@2026-05-22

Confidence: high

 —
Feature Medium

Make XSearch capability model-agnostic via subagent fallback by @colesmcintosh.

Make XSearch capability model-agnostic via subagent fallback by @colesmcintosh.

Source: llm_adapter@2026-05-22

Confidence: high

 —
Feature Medium

Support top_k model setting in GoogleModel, AnthropicModel, and CohereModel by @madanlalit.

Support top_k model setting in GoogleModel, AnthropicModel, and CohereModel by @madanlalit.

Source: llm_adapter@2026-05-22

Confidence: high

 —
Dependency Medium

Bump genai-prices minimum version by @adtyavrdhn.

Bump genai-prices minimum version by @adtyavrdhn.

Source: llm_adapter@2026-05-22

Confidence: low

 —
Bugfix Medium

Fix AnthropicModel dropping code execution tool results when continuing a conversation by @2830500285.

Fix AnthropicModel dropping code execution tool results when continuing a conversation by @2830500285.

Source: llm_adapter@2026-05-22

Confidence: high

 —
Bugfix Medium

Don't drop single BaseModel tool arguments sent in wrapped form when the model uses default extra='ignore' by @macayu17.

Don't drop single BaseModel tool arguments sent in wrapped form when the model uses default extra='ignore' by @macayu17.

Source: llm_adapter@2026-05-22

Confidence: high

 —
Bugfix Medium

Preserve Xai tool result IDs and require xai-sdk 1.6.0 by @FU-max-boop.

Preserve Xai tool result IDs and require xai-sdk 1.6.0 by @FU-max-boop.

Source: llm_adapter@2026-05-22

Confidence: high

 —
Bugfix Medium

Prevent silent loss of thinking for Claude Sonnet 4.6 and Opus 4.6 on Bedrock; add adaptive thinking support.

Prevent silent loss of thinking for Claude Sonnet 4.6 and Opus 4.6 on Bedrock; add adaptive thinking support.

Source: granite4.1:30b@2026-05-22-audit

Confidence: high

 —
Bugfix Medium

Fix thinking being silently dropped for Claude Sonnet 4.6 and Opus 4.6 on Bedrock, and support adaptive thinking efforts by @sarth6.

Fix thinking being silently dropped for Claude Sonnet 4.6 and Opus 4.6 on Bedrock, and support adaptive thinking efforts by @sarth6.

Source: llm_adapter@2026-05-22

Confidence: low

 —
Full changelog

What's Changed

🚀 Features

  • Add pending message queue (ctx.enqueue / agent_run.enqueue) by @DouweM in https://github.com/pydantic/pydantic-ai/pull/4980
  • Add support for MCP background tasks (SEP-1686) by @Alex-Resch in https://github.com/pydantic/pydantic-ai/pull/5192
  • Make XSearch capability model-agnostic via subagent fallback by @colesmcintosh in https://github.com/pydantic/pydantic-ai/pull/5120
  • Support top_k model setting in GoogleModel, AnthropicModel, and CohereModel by @madanlalit in https://github.com/pydantic/pydantic-ai/pull/5558

🐛 Bug Fixes

  • Fix AnthropicModel dropping code execution tool results when continuing a conversation by @2830500285 in https://github.com/pydantic/pydantic-ai/pull/5568
  • Fix thinking being silently dropped for Claude Sonnet 4.6 and Opus 4.6 on Bedrock, and support adaptive thinking efforts by @sarth6 in https://github.com/pydantic/pydantic-ai/pull/5326
  • Don't drop single BaseModel tool arguments sent in wrapped form when the model uses default extra='ignore' by @macayu17 in https://github.com/pydantic/pydantic-ai/pull/5560
  • Preserve Xai tool result IDs and require xai-sdk 1.6.0 by @FU-max-boop in https://github.com/pydantic/pydantic-ai/pull/5355

📦 Dependencies

  • Bump genai-prices minimum version by @adtyavrdhn in https://github.com/pydantic/pydantic-ai/pull/5565

New Contributors

  • @FU-max-boop made their first contribution in https://github.com/pydantic/pydantic-ai/pull/5355
  • @putramkti made their first contribution in https://github.com/pydantic/pydantic-ai/pull/5559
  • @2830500285 made their first contribution in https://github.com/pydantic/pydantic-ai/pull/5568
  • @macayu17 made their first contribution in https://github.com/pydantic/pydantic-ai/pull/5560

Full Changelog: https://github.com/pydantic/pydantic-ai/compare/v1.100.0...v1.101.0

View diff on GitHub

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Share on X Share on Bluesky

Track pydantic-ai

Get notified when new releases ship.

Sign up free

About pydantic-ai

AI Agent Framework, the Pydantic way

All releases →

Related context

Earlier breaking changes

  • v1.95.0 rename ‘built-in tools’ to ‘native tools’, deprecate old fields, register via capabilities=[NativeTool(...)]

Beta — feedback welcome: [email protected]