# Braintrust review

> Braintrust is a platform for evaluating and shipping AI products, with JavaScript/TypeScript SDKs for logging, tracing, evals, and a CLI.

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

## TL;DR

Braintrust is an AI observability platform for tracing production, running evals, and catching regressions before they reach users. This repository holds its JavaScript/TypeScript SDKs and integrations, aimed at teams building AI agents and products. It pairs code-defined evals with production tracing, plus integrations for OpenTelemetry, Temporal, and the browser.

## Braintrust at a glance

- **Category:** [Evals and observability](https://index.devtoolwatch.com/categories/evals-and-observability/)
- **Maintained by:** Braintrust (organization, [@braintrustdata](https://github.com/braintrustdata) on GitHub)
- **License:** Apache 2.0
- **Language:** TypeScript, JavaScript, Makefile
- **Repository:** [github.com/braintrustdata/braintrust-sdk-javascript](https://github.com/braintrustdata/braintrust-sdk-javascript)
- **Website:** [www.braintrust.dev](https://www.braintrust.dev)
- **npm package:** [braintrust](https://www.npmjs.com/package/braintrust) 3.39.0
- **PyPI package:** [braintrust](https://pypi.org/project/braintrust/) 0.48.0
- **Latest release:** [braintrust@3.39.0](https://github.com/braintrustdata/braintrust-sdk-javascript/releases/tag/braintrust%403.39.0) on GitHub, Oct 21, 2026
- **First published:** Jul 17, 2023

## What Braintrust does

- Braintrust defines evals in code with a dataset, a task function, and scorers such as LevenshteinScorer from autoevals.
- Braintrust traces prompts, responses, and tool calls, and tracks latency, cost, and quality in production.
- Braintrust provides an OpenTelemetry span processor and compatibility helpers, including support for span customizers.
- Braintrust offers a Temporal client/worker plugin and workflow interceptors for tracing workflows.
- Braintrust adds application logging and console capture, plus batch tracing for the AI SDK.
- Braintrust turns production traces into eval datasets so regression tests come from real failures.

## How to install Braintrust

**With npm**

```
npm install braintrust autoevals
```

### Requirements

- Node.js with npm to install the JavaScript/TypeScript SDK.
- For braintrust/custom-views, your own @types/react for React 18 or newer.

### First steps

1. Install the braintrust and autoevals packages with npm.
2. Create tutorial.eval.ts that calls Eval with data, a task, and a scorer like LevenshteinScorer.
3. Run the eval with your API key: BRAINTRUST_API_KEY=<YOUR_API_KEY> npx braintrust eval tutorial.eval.ts

## Who Braintrust is for

- TypeScript teams that want to write and run LLM evals as code.
- Teams tracing AI agents in production and inspecting tool calls.
- Projects already using OpenTelemetry or Temporal that need AI tracing.

### Limitations

- Braintrust's SDK sends data to the Braintrust platform and needs an API key to run evals.
- Braintrust's @braintrust/langchain-js and @braintrust/openai-agents packages are deprecated in favor of integrations in the main braintrust package.
- Braintrust keeps its Python SDK in a separate repository, so this repo covers JavaScript/TypeScript only.

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

- Board rank: #282 overall, #15 in Evals and observability, momentum score 34
- Downloads that week: 4.2M (+10% vs 4-week average)
- New GitHub stars per week (last 12 weeks): Aug 9: 2; Aug 16: 0; Aug 23: 0; Aug 30: 0; Sep 6: 0; Sep 13: 0; Sep 20: 1; Sep 27: 0; Oct 4: 1; Oct 11: 0; Oct 18: 0; Oct 25: 0

## Braintrust 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 |

## Braintrust FAQ

### Is Braintrust free and open source?

Yes. Braintrust's source code is on GitHub (braintrustdata/braintrust-sdk-javascript) under the Apache 2.0 license, so you can use it for free. Its website also links to a pricing page (https://www.braintrust.dev/pricing) for any paid plans.

### How do I install Braintrust?

Run `npm install braintrust autoevals`. The install section above has the other options from the project's README.

### Is Braintrust still maintained?

Yes. The braintrustdata/braintrust-sdk-javascript repository was last updated on Oct 21, 2026, and the latest release, braintrust@3.39.0 on GitHub, came out on Oct 21, 2026.

### How popular is Braintrust?

Braintrust has 29 GitHub stars as of Nov 1, 2026. It gained 0 new stars in the week of Oct 25–31, 2026, against a 12-week average of 0 a week. Its main package had about 4.2M downloads that week, +10% against its own 4-week average. That made it #15 among LLM evals and observability tools on the Devtoolwatch Board.

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

Braintrust and Langfuse are both tracked by Devtoolwatch as LLM evals and observability tools. Braintrust has 29 GitHub stars and Langfuse has 36.8k. In the week of Oct 25–31, 2026, Braintrust's main package had 4.2M 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/)

### Braintrust vs LangChain: how do they compare?

Braintrust and LangChain are both tracked by Devtoolwatch. Braintrust has 29 GitHub stars and LangChain has 148.8k. In the week of Oct 25–31, 2026, Braintrust's main package had 4.2M downloads and LangChain's had 41.1M. [LangChain review](https://index.devtoolwatch.com/tools/langchain/)

### What are the best alternatives to Braintrust?

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.

### Does Braintrust work with Python?

Braintrust has a Python SDK on PyPI that requires Python 3.10 or newer. Its code lives in a separate Braintrust Python SDK repository, split out of this JavaScript repo.

### What is the bt CLI in Braintrust?

Braintrust installs the bt CLI alongside the braintrust package. You run it with npx bt or pnpm exec bt.

### How does Braintrust score eval outputs?

Braintrust scores outputs with LLMs, code, or humans. In code, you pass scorers such as LevenshteinScorer from the autoevals package to Eval.

### Can Braintrust cancel long-running evals?

Braintrust supports cooperative cancellation since 3.37.0. On timeout or abort, it stops new trials, aborts hooks.signal, and waits for in-flight tasks and scorers to settle.

## Sources

1. [braintrustdata/braintrust-sdk-javascript on GitHub](https://github.com/braintrustdata/braintrust-sdk-javascript)
2. [Braintrust README](https://github.com/braintrustdata/braintrust-sdk-javascript/blob/main/README.md)
3. [braintrust on npm](https://www.npmjs.com/package/braintrust)
4. [braintrust on PyPI](https://pypi.org/project/braintrust/)
5. [Braintrust website](https://www.braintrust.dev)
6. [Release braintrust@3.39.0 (Oct 21, 2026)](https://github.com/braintrustdata/braintrust-sdk-javascript/releases/tag/braintrust%403.39.0)
7. [Release @braintrust/otel@1.1.0 (Oct 19, 2026)](https://github.com/braintrustdata/braintrust-sdk-javascript/releases/tag/%40braintrust/otel%401.1.0)
8. [Release braintrust@3.38.0 (Oct 19, 2026)](https://github.com/braintrustdata/braintrust-sdk-javascript/releases/tag/braintrust%403.38.0)
9. [Hacker News: The rise of async AI programming](https://news.ycombinator.com/item?id=45210693)
10. [Hacker News: Llama-3.1 supports tool calls via prompting](https://news.ycombinator.com/item?id=41081460)
11. [Hacker News: Brainstore: A full-text search database on Obj Store written in Rust](https://news.ycombinator.com/item?id=43244636)
12. [Devtoolwatch Board, the week of Oct 25–31, 2026](https://index.devtoolwatch.com/)