23 stories
How dynamic shredding makes the vast majority of queries in Logfire a thousand times faster, with early adopter customers reporting more than 1000x improvement on query performance.
Pydantic v2.14 adds Python 3.15 support, with lazy imports and TypeForm, stabilizes the MISSING sentinel, and speeds up schema generation.
Claude Agent SDK puts Claude Code inside your application. If your product is not a coding agent, that convenience comes with architectural decisions you did not make.
Agents are still software, but traditional monitoring can't see what they actually do. A practitioner's case for treating observability as a first-class layer, and how we do it with Logfire.
Connect Pydantic Logfire's MCP server to Cursor, Claude Code, or Codex and let your coding agent query real production traces, then build the dashboard and the alert that catch the next one.
Pydantic AI can run an agent on TypeSafe's Jev, a model that answers typed questions instead of writing text. Pydantic Logfire now shows each question, answer, and probability in Live view.
Bring your TypeSafe key to Pydantic AI Gateway and call Jev through the System One API, with metered cost, guardrails, and the same gateway key you use for every other model.
Use Jev for categorical judgments, probabilities, and numeric rubric scores in Pydantic Evals, then track the results in Logfire without a per-score fee.
We overhauled the look and feel of Logfire dashboards. Here's the design thinking, and what's next.
Image generation is now part of Pydantic AI. Here's how to use it, why we added it, and the contributor conversations that helped it happen.
AWSLambdaDurability checkpoints every model request, tool call, and MCP call as an AWS Lambda durable step. An invocation that times out or gets retried resumes from the last completed step instead of paying twice for work the agent already did.
Frontier models are changing the words in our codebases, and the change spreads from one model to the next. Evidence from GitHub, and a Pydantic AI Capability that pushes back.
You.com ships as two capabilities in Pydantic AI Harness. YouSearch gives an agent web search with excerpts and page reads, YouResearch gives it cited answers and multi-step research, and both hand your application structured citations it can render without parsing text.
A buyer's comparison of seven LLM evaluation tools: what each one meters, whether it scores curated datasets or live production traffic, what its license actually permits, and how much of your setup survives a switch.
A practical guide to Langfuse alternatives in 2026: what each one is licensed under, what its meter counts, who it suits, and when staying on Langfuse is the better call.
An honest comparison of seven prompt management tools in 2026: who each one is for, how versioning, promotion, and rollout actually differ, and when prompts in your repository are still the right call.
A practical guide to OpenTelemetry backends in 2026: who each one is actually for, how each one meters usage, and what changes when your app includes LLMs.
Realtime speech-to-speech support in Pydantic AI: the same agent, tools, and message history you already have, now over a live voice call. Provider-agnostic across OpenAI, Azure OpenAI, Gemini Live, and xAI, with browser WebRTC, server-side tools, gateway routing, and Logfire tracing.
Showing 18 of 23 stories
Load more storiesGet the top AI news and tools straight to your inbox every week.
Weekly · AI news
Join 15,000+ AI enthusiasts
