# Braintrust vs Kernel: who uses which

On usedby, 44 companies use Braintrust and 17 use Kernel. 1 company uses both, including Vercel. Counts include every proof level, job-post signals included; each company keeps its proof level on each side. Latest check 2026-10-25: 61 of 61 public lines checked.

## Side by side

|  | [Braintrust](https://www.usedby.ai/tools/braintrust.md) | [Kernel](https://www.usedby.ai/tools/kernel.md) |
|---|---|---|
| Public customers | 44 | 17 |
| Confirmed | 13 | 8 |
| Observed | 30 | 9 |
| Signals (job posts) | 1 | 0 |
| Top industries of customers (every proof level) | DevTools & Infrastructure (12), Artificial Intelligence (5), Fintech & Payments (5) | DevTools & Infrastructure (2), Artificial Intelligence (1), Biotech & Pharma (1) |
| Category | LLM observability & evals | Agent frameworks & AI gateways |
| Security | SOC 2 Type II ([trust.braintrust.dev](https://trust.braintrust.dev/), checked 2026-10-22) | — |
| Data processing terms | A data processing agreement is available. ([braintrust.dev](https://www.braintrust.dev/legal/dpa), checked 2026-10-22) | — |
| Free trial | Free trial or free plan available. ([braintrust.dev](https://www.braintrust.dev/pricing), checked 2026-10-22) | — |
| Integrations | Vercel AI SDK, Claude Agent SDK, OpenAI Agents SDK, LlamaIndex, CrewAI, LangChain, LangGraph, OpenTelemetry, Mastra, Pydantic AI, DSPy, Pytest ([braintrust.dev](https://www.braintrust.dev/docs/integrations), checked 2026-10-22) | — |
| Pricing | Starter: $0 / month; Pro: $249 / month; Enterprise: contact sales ([braintrust.dev](https://www.braintrust.dev/pricing), checked 2026-10-22) | — |

## Companies that use both

Every proof level counts; the proof level of each side is shown, strongest proofs first. The first public proof is not an adoption date.

| Company | Proof, Braintrust | Proof, Kernel | First public proof, Braintrust |
|---|---|---|---|
| [Vercel](https://www.usedby.ai/companies/vercel.md) | Observed | Observed | On or before April 2025 |

## Published results

Reviewed stories only; usedby did not measure the figures.

- Braintrust: [Browserbase](https://www.usedby.ai/case-studies/browserbase-braintrust.md): **100,000s**: “100,000s of public eval dashboard views” (captured by usedby 2026-10-09)
- Braintrust: [Rex](https://www.usedby.ai/case-studies/rex-braintrust.md): **30-min**: “Continuous signal, 30-min fix cycles” (captured by usedby 2026-10-06)
- Kernel: [Stably](https://www.usedby.ai/case-studies/stably-kernel.md): **333x**: “cutting browser cold starts by 333x” (published 2026-05-26)
- Kernel: [Flint](https://www.usedby.ai/case-studies/flint-kernel.md): **167x**: “167x faster cold starts” (published 2026-03-30)

---
Source: https://www.usedby.ai/compare/braintrust-vs-kernel · How we check: https://www.usedby.ai/methodology