Baserun Review (2026)
A YC-backed LLM testing and observability platform, acquired by LlamaIndex. Its own X account is marked "(Acquired)" and no standalone product remains to evaluate.
Rating
Starting Price
Unavailable
Free Plan
No
SDKs & Frameworks
2
Deployment
1
Best For
Nobody as a standalone product. If you are a LlamaIndex user, the relevant question is what of this capability now appears there.
Last Updated:
10 Things You Should Know About Baserun
- 1 Acquired by LlamaIndex
- 2 Baserun's own X account describes the company and is marked "(Acquired)"
- 3 Founded in 2023 in San Francisco by Adam Ginzberg, Erik Peterson and Effy Zhang
- 4 Raised roughly $500K in funding through Y Combinator
- 5 The platform offered a prompt playground, end-to-end tests and a visualization dashboard
- 6 No acquisition announcement date or terms could be found in public sources
Pros & Cons
Pros
- ✓ The acquirer is an active, well-known project rather than a company winding the technology down
- ✓ The core idea - a single loop from spotting an issue to evaluating the fix - was a sound framing
Cons
- ✕ Acquired, with no standalone product available to evaluate or buy
- ✕ No acquisition date or terms disclosed in any source we could find
- ✕ No public migration guidance for former users that we could locate
- ✕ Only $500K raised, so it was a small team and the acquisition was likely a talent and technology deal rather than a product continuation
- ✕ Anyone who built on it needs to establish directly what happens to their data and integrations
Features
Baserun was acquired by LlamaIndex
There is no standalone product here to evaluate.
Baserun’s own X account describes the platform and is marked “(Acquired)”. The acquirer is LlamaIndex.
We could not find an announcement date, deal terms, press release or migration guidance in any source. That absence is itself informative - small acquisitions frequently pass without formal announcement, and this one appears to have done so.
How quietly this happened
Worth dwelling on, because it is the useful part of this page.
This site now documents a lot of consolidation: Helicone into Mintlify and then maintenance mode, Humanloop acqui-hired by Anthropic and shut down, Traceloop into ServiceNow, TruEra into Snowflake, promptfoo into OpenAI, W&B into CoreWeave, Literal AI, WhyLabs and Gentrace shut down outright.
Every one of those left a visible trace - a press release, a banner on the vendor’s site, a farewell post, a shutdown notice.
Baserun’s acquisition is visible as a parenthetical in a social media bio.
The lesson for anyone maintaining a shortlist: the absence of shutdown news is not evidence that a small vendor is still independent. Directory listings and comparison articles will carry a company for years after it stops existing as a standalone entity. The check that catches this is looking at the company’s own channels - site banner, X bio, changelog, commit history - not at anyone’s comparison table, including this one.
What it was
Founded in 2023 in San Francisco by Adam Ginzberg, Erik Peterson and Effy Zhang, having raised roughly $500K through Y Combinator.
The platform offered a prompt playground, end-to-end tests, and a visualization dashboard, positioned around streamlining the development cycle from identifying an issue through to evaluating the solution.
That framing was sound and is worth carrying forward as a requirement. The loop from “something is wrong in production” to “here is a test that proves it is fixed” is the thing most tools in this category handle badly, and it is a reasonable criterion to judge a replacement on.
What the funding tells you
Roughly $500K is a small raise, meaning a small team and limited runway by this category’s standards.
Acquisitions at that scale are typically about the people and the technology rather than a commitment to operate the acquired product. That reading fits what we observed: no announcement, no migration documentation, no continued marketing.
The better news is the acquirer. LlamaIndex is an active, widely used project rather than a company absorbing a competitor to shut it down - so the technology has somewhere to go. Whether Baserun’s specific features surface as LlamaIndex functionality, get rebuilt, or are simply absorbed as team expertise is not documented anywhere we could find. If that matters to you, ask LlamaIndex.
If you are still running it
Contact LlamaIndex directly, and do not wait for a notice.
We found no public sunset timeline and no statement about existing accounts or data. That does not mean nothing was communicated - it may have gone to customers privately - but you cannot plan from public information. Get written answers on how long the service runs, what your export options are, and whether data is retained or deleted. Take local copies of anything you would not want to lose.
What to use instead
- Agenta or Latitude for the prompt playground and iteration workflow. Both MIT, both self-hostable.
- promptfoo or DeepEval for end-to-end testing in CI.
- Langfuse for observability with evaluation attached.
All actively maintained with real communities - the specific thing an acquired product can no longer offer.
Should you use it?
No. There is nothing to adopt.
Bottom line: a small, sensibly designed platform absorbed into LlamaIndex with no public announcement. The product is gone; the workflow idea is worth keeping. The broader lesson is about how to check, because this is the quietest disappearance we have documented and nothing in the usual comparison sources reflects it.
Acquisition confirmed from the company’s own social media profile and third-party company databases on 3 August 2026. No announcement date, deal terms or migration guidance could be found in any public source. This is a researched directory entry - we have not instrumented this platform, and given the acquisition, we will not.
Pricing Plans
Standalone platform
Unavailable
- Company acquired by LlamaIndex
- No standalone product to purchase
- Acquisition date and terms not disclosed publicly
SDKs & Frameworks
Deployment
Eval Methods
Status
Our Verdict
Baserun no longer exists as something you can buy. It was acquired by LlamaIndex, and its own X account now carries "(Acquired)" in the description. We could not find an announcement date, deal terms or public migration guidance in any source, which is worth noting for anyone still running it - you will need to contact LlamaIndex directly rather than relying on public information. The company was small - founded in 2023 in San Francisco by Adam Ginzberg, Erik Peterson and Effy Zhang, having raised around $500K through Y Combinator. At that size an acquisition is usually a team and technology deal rather than a commitment to keep the standalone product running, and the absence of any migration documentation is consistent with that reading. The underlying idea was sound - one workflow spanning issue identification, prompt iteration and evaluation of the fix - and that framing is worth carrying into whatever you choose instead. The product itself is not an option.
Similar Tools
Vellum
Nobody evaluating prompt management or LLM development tooling. If you want a personal AI assistant, the current product is a different conversation entirely.
Agenta
Teams that want one self-hosted MIT-licensed platform covering prompts, evaluation and observability, and who value environment promotion and branching over best-in-class depth in any single area.
Freeplay
Cross-functional product teams where non-engineers own prompt iteration and need to ship changes without a deploy, and who can work with enterprise procurement rather than self-serve pricing.
Latitude
Teams that want self-hosted open-source prompt management with reusable components, and who care about closing the loop from production behaviour back to prompt changes.
Frequently Asked Questions
What happened to Baserun?
It was acquired by LlamaIndex. Baserun's own X account describes it as a collaborative platform for engineers and product experts to build, monitor and improve AI, and is marked "(Acquired)". We could not find an announcement date, deal terms or a press release in any source, which is itself informative - small acquisitions frequently pass without formal announcement. If you need the date for contractual or audit reasons, your own account records and direct contact with LlamaIndex are your only reliable options.
I still use Baserun. What should I do?
Contact LlamaIndex directly, and do not wait. We found no public migration guidance, no sunset timeline and no statement about what happens to existing accounts or data. That absence does not mean nothing was communicated - it may have gone to customers privately - but it does mean you cannot plan from public information. Establish in writing how long the service will run, what your export options are, and whether your data will be retained or deleted. In the meantime, take local copies of anything you would not want to lose.
Does the capability live on inside LlamaIndex?
We could not establish this and are not going to guess. LlamaIndex is an active, widely used framework, so an acquisition there is a better outcome than an outright shutdown - the technology has somewhere to go. But whether Baserun's specific features appear as LlamaIndex functionality, get rebuilt, or are quietly absorbed as team expertise is not documented anywhere we could find. If this matters to your decision, ask LlamaIndex.
Why does a $500K raise matter here?
It sets expectations for what the acquisition probably was. Baserun raised roughly $500K through Y Combinator, which means a small team and a short runway by the standards of this category. Acquisitions at that scale are usually about the people and the technology rather than a commitment to operate the acquired product. That reading is consistent with what we observed - no announcement, no migration documentation, no continued marketing. It is not a criticism of anyone involved, but it is the realistic frame for anyone hoping the standalone platform continues.
What should I use instead?
For the prompt playground and iteration workflow, Agenta or Latitude are the strongest open-source options and both are MIT with self-hosting. For end-to-end testing of LLM applications in CI, promptfoo and DeepEval are built for that. For observability with evaluation attached, Langfuse is the safest default. All are actively maintained with real communities, which is the specific thing an acquired product can no longer promise.
Is this the same pattern as other acquisitions in this category?
It is the quietest version of it. This site now documents Helicone into Mintlify and then maintenance mode, Humanloop acqui-hired by Anthropic and shut down, Traceloop into ServiceNow, TruEra into Snowflake, promptfoo into OpenAI, and W&B into CoreWeave. Those all had announcements. Baserun's is visible only as a parenthetical on a social media profile. For anyone maintaining a shortlist, that is the useful lesson - the absence of shutdown news is not evidence a small vendor is still independent, and the check that catches this is looking at the company's own channels rather than comparison articles.