# Logfire review

> Logfire is an observability platform from the Pydantic team for LLM apps, agents, and Python services, built on OpenTelemetry.

Devtoolwatch Board, the week of Oct 25–31, 2026. Canonical page: https://index.devtoolwatch.com/tools/logfire/

## TL;DR

Logfire is an observability platform from the team behind Pydantic Validation, built as an opinionated wrapper around OpenTelemetry. It is for developers running LLM apps, agents, and Python services in production. What sets it apart: you query traces, logs, metrics, and eval results with standard SQL, and the Python SDK gives detailed visibility into Python code.

## Logfire at a glance

- **Category:** [Evals and observability](https://index.devtoolwatch.com/categories/evals-and-observability/)
- **Maintained by:** Pydantic (organization, [@pydantic](https://github.com/pydantic) on GitHub)
- **License:** MIT
- **Language:** Python, JavaScript, Makefile
- **Repository:** [github.com/pydantic/logfire](https://github.com/pydantic/logfire)
- **Website:** [pydantic.dev/logfire](https://pydantic.dev/logfire/)
- **npm package:** [@pydantic/logfire-node](https://www.npmjs.com/package/@pydantic/logfire-node) 0.18.30
- **PyPI package:** [logfire](https://pypi.org/project/logfire/) 5.2.0
- **Latest release:** [@pydantic/logfire-node 0.18.30](https://www.npmjs.com/package/@pydantic/logfire-node) on npm, Oct 21, 2026
- **First published:** Apr 23, 2024

## What Logfire does

- Logfire supports all OpenTelemetry signals: traces, metrics, and logs.
- Logfire lets you query your data with standard SQL, including from existing BI tools and database querying libraries.
- Logfire shows rich displays of Python objects, event-loop telemetry, and profiling of Python code and database queries.
- Logfire traces LLM calls and costs, and stores eval results and prompt versions in the same table as traces.
- Logfire auto-instruments popular packages such as FastAPI, so you can skip most manual instrumentation.
- Logfire gives analytics on data flowing through Pydantic Validation models.

### How it works

Logfire has open-source SDKs for Python, TypeScript, and Rust that send telemetry to the closed-source Logfire platform, which holds the UI and backend. Other languages connect through standard OpenTelemetry. You call logfire.configure() in your code, then add spans and logs or instrument libraries. The SDKs can also export to any OTel-compatible backend. Model calls routed through the AI Gateway land in Logfire without an SDK.

## How to install Logfire

**With pip (SDK and CLI)**

```
pip install logfire
```

**With pip (SDK only)**

```
pip install logfire-sdk
```

**Set up with your coding agent**

```
uvx logfire-cli wizard
```

### Requirements

- Python 3.10 or later for the Python SDK.
- A Logfire account, authenticated with logfire auth, to send data to the hosted platform.
- Self-hosting the platform requires purchasing an enterprise license.

### First steps

1. Install Logfire with pip install logfire, which adds both the Python SDK and the CLI.
2. Authenticate by running logfire auth.
3. Call logfire.configure() in your code, then send logs with logfire.info() or wrap work in logfire.span().
4. Instrument frameworks instead of tracing manually, for example with logfire.instrument_fastapi(app).

## Who Logfire is for

- Teams running LLM apps and agents in production that need tracing, evals, and cost tracking.
- Python developers, especially those using FastAPI, Pydantic, or Pydantic AI.
- Teams that want to query telemetry with SQL or from existing BI tools.
- Polyglot teams already using OpenTelemetry across languages.

### Limitations

- Logfire's platform, meaning the UI and backend, is closed source. Only the SDKs are open source.
- Logfire can only be self-hosted with a purchased enterprise license.
- Logfire has native SDKs only for Python, TypeScript, and Rust. Other languages rely on standard OpenTelemetry.
- Logfire's zstd compression support is experimental as of v5.2.0.

## Usage data (the week of Oct 25–31, 2026)

- Board rank: #226 overall, #11 in Evals and observability, momentum score 41
- Downloads that week: 4.6M (+16% vs 4-week average)
- New GitHub stars per week (last 12 weeks): Aug 9: 13; Aug 16: 6; Aug 23: 14; Aug 30: 11; Sep 6: 17; Sep 13: 7; Sep 20: 15; Sep 27: 13; Oct 4: 5; Oct 11: 12; Oct 18: 5; Oct 25: 9

## Logfire alternatives

| Rank | Tool | Stars | Downloads/wk |
|---|---|---|---|
| #1 | [SkillSpector](https://index.devtoolwatch.com/tools/nvidia-skillspector/) | 23.7k | – |
| #2 | [Judgeval](https://index.devtoolwatch.com/tools/judgmentlabs-judgeval/) | 1.1k | 80.2k |
| #3 | [A.I.G (AI-Infra-Guard)](https://index.devtoolwatch.com/tools/tencent-ai-infra-guard/) | 7.2k | – |
| #4 | [Langfuse](https://index.devtoolwatch.com/tools/langfuse/) | 36.8k | 9.8M |
| #5 | [Opik](https://index.devtoolwatch.com/tools/opik/) | 23.1k | 709.1k |
| #6 | [promptfoo](https://index.devtoolwatch.com/tools/promptfoo/) | 26.7k | 1.2M |

## Logfire FAQ

### Is Logfire free and open source?

Yes. Logfire's source code is on GitHub (pydantic/logfire) under the MIT license, so you can use it for free. Its website also links to a pricing page (https://pydantic.dev/pricing) for any paid plans.

### How do I install Logfire?

Run `pip install logfire`. The install section above has the other options from the project's README.

### Is Logfire still maintained?

Yes. The pydantic/logfire repository was last updated on Oct 22, 2026, and the latest release, @pydantic/logfire-node 0.18.30 on npm, came out on Oct 21, 2026.

### How popular is Logfire?

Logfire has 4.5k GitHub stars as of Nov 1, 2026. It gained 9 new stars in the week of Oct 25–31, 2026, against a 12-week average of 11 a week. Its main package had about 4.6M downloads that week, +16% against its own 4-week average. That made it #11 among LLM evals and observability tools on the Devtoolwatch Board.

### Logfire vs Langfuse: how do they compare?

Logfire and Langfuse are both tracked by Devtoolwatch as LLM evals and observability tools. Logfire has 4.5k GitHub stars and Langfuse has 36.8k. In the week of Oct 25–31, 2026, Logfire's main package had 4.6M downloads and Langfuse's had 9.8M. Langfuse is an open source LLM engineering platform for tracing, evaluating, prompt management, and debugging LLM applications and agents. [Langfuse review](https://index.devtoolwatch.com/tools/langfuse/)

### What are the best alternatives to Logfire?

Other LLM evals and observability tools on the Devtoolwatch Board in the week of Oct 25–31, 2026, by momentum: SkillSpector (#1), Judgeval (#2), A.I.G (AI-Infra-Guard) (#3), Langfuse (#4), Opik (#5). The alternatives table above has their numbers.

### What is Pydantic Logfire used for?

Logfire is used for observability of LLM apps, agents, APIs, and general applications. It covers distributed tracing, logs, metrics, evals, and cost tracking. You can query all of this data with SQL.

### Is Logfire open source?

Logfire is partly open source. The SDKs for Python, TypeScript, and Rust are open source. The platform, meaning the UI and backend, is closed source.

### Can I use the Logfire SDK with an alternative backend?

Yes. The open-source Logfire SDKs can export data to any OTel-compatible backend, so you are not tied to the Logfire platform.

### How do I use Pydantic Logfire in Python?

Install Logfire with pip install logfire and run logfire auth. Then call logfire.configure() in your code and use logfire.info() or logfire.span(). You can also instrument a framework, for example with logfire.instrument_fastapi(app).

### Is there a free tier for Logfire?

Yes. The Logfire homepage lists 10 million spans free every month, with no credit card needed.

## Sources

1. [pydantic/logfire on GitHub](https://github.com/pydantic/logfire)
2. [Logfire README](https://github.com/pydantic/logfire/blob/main/README.md)
3. [@pydantic/logfire-node on npm](https://www.npmjs.com/package/@pydantic/logfire-node)
4. [logfire on PyPI](https://pypi.org/project/logfire/)
5. [Logfire website](https://pydantic.dev/logfire/)
6. [Release v5.2.0 (Oct 13, 2026)](https://github.com/pydantic/logfire/releases/tag/v5.2.0)
7. [Release v6.0.0b8 (Oct 10, 2026)](https://github.com/pydantic/logfire/releases/tag/v6.0.0b8)
8. [Release v6.0.0b7 (Sep 25, 2026)](https://github.com/pydantic/logfire/releases/tag/v6.0.0b7)
9. [Hacker News: The human-in-the-loop is tired](https://news.ycombinator.com/item?id=48942000)
10. [Hacker News: Pydantic Logfire](https://news.ycombinator.com/item?id=40212490)
11. [Hacker News: Pydantic V2 rewritten in Rust is 5-50x faster than Pydantic V1](https://news.ycombinator.com/item?id=35490449)
12. [Devtoolwatch Board, the week of Oct 25–31, 2026](https://index.devtoolwatch.com/)