LLM observability & evals tools used by DevTools & Infrastructure companies
Each line links its public proof and the date usedby last checked it.
17
DevTools & Infrastructure companies use LLM observability & evals tools
The most used: Braintrust (12), LangSmith (3), and Arize AI (2).Data as of Oct 25, 2026
Ranking
Tools ranked by the DevTools & Infrastructure companies that use them. Ties go to the stronger proof.
Who uses each tool
Strongest proof first. A job post is a signal, the weakest proof.
1
Braintrust12 companies · Checked Oct 25, 2026126 confirmed 6 observed
DropboxChecked Oct 25, 2026
RetoolChecked Oct 25, 2026
BoxChecked Oct 25, 2026
BrowserbaseChecked Oct 25, 2026
BubbleChecked Oct 25, 2026GraphiteChecked Oct 25, 2026
AirtableChecked Oct 25, 2026
VercelChecked Oct 25, 2026
ReplitChecked Oct 25, 2026
Augment CodeChecked Oct 25, 2026
Show 2 more companies
GreptileChecked Oct 25, 2026
SuperhumanChecked Oct 25, 2026
2
LangSmith3 companies · Checked Oct 23, 202630 confirmed 3 observed3
Arize AI2 companies · Checked Oct 25, 202624
Comet Opik1 company · Checked Oct 21, 20261
How usedby counts
A company counts once per tool, at its strongest proof: confirmed (a case study, a named testimonial, a press release or the company's own account), declared (the company told usedby), observed (a vendor logo wall or customers page) or signal (a job post). Counts cover the companies on usedby, never the whole market.
No software vendor had to be left out.
Data as of October 25, 2026.
Questions
Which LLM observability & evals tools do DevTools & Infrastructure companies use most?
Braintrust leads: 12 of the 17 DevTools & Infrastructure companies on usedby that use LLM observability & evals tools use it. Next: LangSmith (3), Arize AI (2), and Comet Opik (1). See Braintrust.
How many DevTools & Infrastructure companies use LLM observability & evals tools?
17 DevTools & Infrastructure companies on usedby use at least one of 4 LLM observability & evals tools. It counts the companies usedby tracks.
What proof backs these lines?
Of the 17 companies, each counted at its strongest proof: 7 confirmed, 0 declared, 10 observed and 0 known from a job post.
How does usedby build this ranking?
usedby counts each DevTools & Infrastructure company once per tool, from public proof: case studies, press releases, the company's own site or its declaration to usedby, vendor customer pages and job posts. Every line shows its proof and the date usedby last checked it.