LangWatch
OpenTelemetry-native LLMOps for tracing, evals and agent simulation
Overview
LangWatch is an Amsterdam-built LLMOps platform, Apache-2.0 since May 2026, for tracing LLM calls and agents via OpenTelemetry, running offline and live evaluations, simulating and red-teaming agents, optimising prompts with DSPy and routing models through an AI gateway. Free Developer plan; Growth €29 (US$34) per core seat a month.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 1 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- 50k events a month
- 14-day data access
- 2 users
- 3 scenarios, simulations and custom evals
- 200k events included
- 30-day retention
- Unlimited lite users
- Unlimited simulations, evals and prompts
- Private Slack/Teams support
- Custom SSO, RBAC and retention
- Audit logs and SLAs
- ISO 27001 reports
- Marketplace billing
What you can produce with LangWatch
- OpenTelemetry-native LLM and agent tracing
- Offline and live evaluations
- Agent simulations, red-teaming and voice testing
- Prompt versioning and DSPy optimisation
- AI gateway with virtual keys and budgets
- Apache-2.0 self-hosting
LangWatch has become a genuinely open-source LLMOps platform, with strong agent simulation, evaluation and gateway features and generous self-hosting terms. Its pricing advantage over rivals is argued in its own blog rather than independently, and it is a young, small company. A strong choice for teams wanting open, self-hostable observability and testing; compare event volumes against Growth pricing before scaling.
What we found
LangWatch is an LLMOps platform for monitoring, evaluating and improving AI applications and agents. Built on OpenTelemetry, it traces LLM calls and agent steps from Python, TypeScript and Go SDKs or any OpenTelemetry source, runs evaluations offline and on live traffic, simulates conversations with agent scenarios, red-teams applications, tests voice agents, versions prompts and optimises them with DSPy. An AI gateway issues virtual keys per application, team or person, with budgets and provider fallback, and newer additions include Langy, an automated AI engineer, instant evals and Claude Code usage tracking. In May 2026 LangWatch moved its platform from the source-available BSL licence to Apache 2.0, keeping enterprise modules under a separate licence, and self-hosted installations with Docker and ClickHouse now have no caps on members or projects.
LangWatch is run by LangWatch Reasoning Engine B.V. in Amsterdam. Its co-founders, chief executive Manouk Draisma, formerly a commercial leader at Lightspeed and Biller, and technical chief Rogerio Chaves, who helped build Booking.com's first AI agents, met through Antler. It raised a €1 million pre-seed round led by Passion Capital with Volta Ventures and Antler in February 2025.
The cloud Developer plan is free forever with 50,000 events a month, 14 days of data, two users and three each of scenarios, simulations and custom evaluations. Growth costs €29, or US$34, per core seat a month with 200,000 events included and €5 per extra 100,000, 30-day retention, unlimited lite users and unlimited simulations and evaluations. Enterprise adds hybrid, self-hosted or on-premises deployment, SSO, audit logs, ISO 27001 reports and marketplace billing. LangWatch's own blog argues its event-based pricing is cheaper than per-trace rivals; that comparison is its own.
We could not read independent review sites. LangWatch suits teams wanting an open-source, OpenTelemetry-native alternative to commercial LLM observability tools.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
About this listing
This entry was compiled from publicly available data including LangWatch's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with LangWatch unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to LangWatch directly →
Spotted something out of date? Suggest an update →
Related reading
Cybersecurity
CISA Orders Three-Day Fix for Exploited Collabora Flaw in Nextcloud's Built-In Office Serv...
AI & ML
DeepSeek Open-Sources DeepInfer, the Engine That Serves Its V4.1 Flash Model on NVIDIA and...
AI & ML
Google Stopped Paying for One Kind of Open-Source Bug Report After a Flood of Invalid Ones
More in Code & Dev Tools
