pydantic-ai-absurd
Durable execution of Pydantic AI agents on Postgres via Absurd.
Decision gist · record as of 2026-08-14
Yes, if you run Pydantic AI agents in production and want durability without external infrastructure. The low install friction, permissive license, and active maintenance make it a practical choice for teams already using Postgres. Verify that Postgres schema setup and performance overhead align with your deployment model before committing.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires Postgres instance and supports Python 3.10 or later; agent must be run inside an Absurd task context.
- Low friction: pure Python wheel with only three runtime dependencies (pydantic-ai, absurd-sdk, typing-extensions).
- Active maintenance status and recent release suggest ongoing support.
License · maintenance · safety
MIT (permissive) — MIT license permits commercial use, modification, and distribution with minimal restrictions—suitable for most production deployments.
last release 2026-07-28 (17 days)
0 known vulnerabilities (OSV.dev, 2026-08-14) · 104,826 downloads/mo, #12,735 on PyPI
Alternatives
Verify before relying
pip install pydantic-ai-absurd
from absurd_sdk import AsyncAbsurd
from pydantic_ai import Agent
from pydantic_ai_absurd import AbsurdDurability
absurd = AsyncAbsurd("postgresql://localhost/absurd", queue_name="agents")
agent = Agent("openai:gpt-5.2", capabilities=[AbsurdDurability()])
@absurd.register_task(name="analyse")
async def analyse(params, ctx):
result = await agent.run(params["prompt"])
return {"output": result.output}- Whether Postgres schema is auto-created or requires manual setup.
- Performance overhead of checkpointing on latency-sensitive agent workloads.
- Compatibility with all Pydantic AI model providers and tool types.
What it is and what it does
Pydantic AI Absurd wraps Pydantic AI agents with durable execution backed by Postgres. When an agent runs inside an Absurd task, every model call and tool invocation is checkpointed to the database. If the worker process crashes mid-execution, a new worker can resume from the last completed step without restarting the entire run or re-spending tokens on already-completed API calls.
The package integrates Absurd's task queue and durability layer directly into Pydantic AI's agent runtime, eliminating the need for external infrastructure like Temporal or Redis. You define tasks using decorators, spawn them with parameters, and call agent.run() inside the task—Absurd handles checkpointing and recovery transparently. It's designed for scenarios where agent runs are long, expensive, or prone to interruption.
Use it for
- Long-running analysis agents that call LLMs multiple times and must survive worker restarts without token waste.
- Multi-step agentic workflows with tool calls that need to resume from the last completed step after crashes.
- Production deployments where Postgres is already present and external orchestration services are not available.
- Cost-sensitive agent applications where re-running expensive model calls after a failure is unacceptable.
- Batch processing of agent tasks with guaranteed completion despite transient worker failures.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you run Pydantic AI agents in production and want durability without external infrastructure.
The low install friction, permissive license, and active maintenance make it a practical choice for teams already using Postgres. Verify that Postgres schema setup and performance overhead align with your deployment model before committing.
Install
pydantic-ai-absurd on PyPI
Before you install
Low friction: pure Python wheel with only three runtime dependencies (pydantic-ai, absurd-sdk, typing-extensions). Active maintenance status and recent release suggest ongoing support.
Requires Postgres instance and supports Python 3.10 or later; agent must be run inside an Absurd task context.
License in practice
MIT license permits commercial use, modification, and distribution with minimal restrictions—suitable for most production deployments.
Quickstart
pip install pydantic-ai-absurd
from absurd_sdk import AsyncAbsurd
from pydantic_ai import Agent
from pydantic_ai_absurd import AbsurdDurability
absurd = AsyncAbsurd("postgresql://localhost/absurd", queue_name="agents")
agent = Agent("openai:gpt-5.2", capabilities=[AbsurdDurability()])
@absurd.register_task(name="analyse")
async def analyse(params, ctx):
result = await agent.run(params["prompt"])
return {"output": result.output}Verify before relying
- Whether Postgres schema is auto-created or requires manual setup.
- Performance overhead of checkpointing on latency-sensitive agent workloads.
- Compatibility with all Pydantic AI model providers and tool types.
Package facts
| License | MIT permissive |
| Python support | Supports the current Python release >=3.10 |
| Install friction | Low. Pure-Python wheel |
| Runtime dependencies | 3 packagesabsurd-sdkpydantic-aityping-extensions |
| Maintenance | Actively maintained 17 days since the last release |
| First released | |
| Downloads | 104,826 / month, #12,735 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Development Status :: 3 - AlphaIntended Audience :: DevelopersOperating System :: OS IndependentProgramming Language :: Python :: 3Programming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.13Programming Language :: Python :: 3.14 |
Evidence: pydantic_ai_absurd-0.7.0-py3-none-any.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “pydantic ai fault tolerance”
- pydantic-ai-absurdAdds durable execution to Pydantic AI agents by checkpointing model…
- mistralai-workflowsMistral Workflows is a Python SDK for building AI-powered workflows…
- torchft-nightlyProvides per-step fault tolerance for PyTorch distributed training,…
Give your agent the search over MCP, or paste the wish link into any chat.
More Artificial Intelligence packages
LiteLLM provides a unified Python interface to call 100+ LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, and others) using OpenAI-compatible API format, available as both a Python SDK and a self-hosted AI Gateway proxy server.
Install it if you need to work with multiple LLM providers or want to centralize LLM routing in your organization.
Client library and CLI tool for downloading, uploading, and managing models, datasets, and repositories on the Hugging Face Hub platform.
Install it if you work with Hugging Face Hub models or datasets.
LangChain provides a framework for building agents and LLM-powered applications by composing language models, tools, and memory through a unified API that abstracts over multiple model providers.
hf-xet provides chunk-based deduplication and efficient file transfer for the Hugging Face Hub, enabling faster uploads and downloads of large files with local disk caching.
Tokenizers converts raw text into token sequences for NLP models, with support for training custom vocabularies and using pre-built tokenizers (BPE, WordPiece) optimized for speed via Rust.
Transformers provides a unified framework for loading, fine-tuning, and running state-of-the-art pretrained models across text, vision, audio, video, and multimodal tasks using PyTorch, JAX, or TensorFlow.
Install it if you need to run or train any transformer-based model for NLP, vision, audio, or multimodal tasks.
See also absurd-sdk · dbos · aws-durable-execution-sdk-python · pydantic-ai-todo · restate-sdk · conductor-python · pydantic-ai · agent-framework-durabletask · agent-framework-foundry-hosting · agent-framework-azurefunctions