LangWatch
LLM observability, evaluation and agent-simulation platform, open source and also hosted.
LLM observability and evals Official MCP server Open source
＋ Build a stack with LangWatch Website Official pricing GitHub
Price
Free (open source)
you pay only for the compute it runs on
Free tier
Yes
Developer plan: 50k events/month, 14-day data access, 2...
SDK downloads
90.1K
npm langwatch · PyPI langwatch
How LangWatch compares
LangWatch is one of 18 APIs in the LLM observability and evals category on AgentGid. It is one of 17 that can be started for free. It ships an official MCP server, like 4 others here. Its official SDK is the 6th most downloaded in the category (90.1K a week). See the full table.
Pricing
| Free tier | Developer plan: 50k events/month, 14-day data access, 2 users, no credit card |
|---|---|
| Cheapest paid plan | — |
| Usage price | Growth: EUR 29 per core-seat/month incl. 200k events, then EUR 5 per 100k events (prices shown in EUR only) |
| Good to know | Pricing page lists EUR only, so numeric USD fields are null. Developer free forever; Growth EUR 29/core-seat/month with 30-day retention (extra retention EUR 3/GB); Enterprise custom. Self-hosting with docker compose is supported; repo is Apache-2.0. Unit is 'events' (every LLM call, tool call, retrieval or eval step). |
From the official pricing page, checked Oct 20, 2026. Prices change: confirm before you buy.
Use it from an agent
| MCP server | Official: @langwatch/mcp-server |
|---|---|
| npm | npm install langwatch |
| PyPI | pip install langwatch |
| GitHub | langwatch/langwatch · 4,988 stars · last push Oct 23, 2026 |
Alternatives to LangWatch
| API | Free | Price | SDK downloads / wk | GitHub stars | MCP |
|---|---|---|---|---|---|
| LangSmith Tracing, evaluation and monitoring of LLM apps and agents, including prompt and dataset management. |
Developer plan: 1 seat, up to 5k base traces... | Free plan · paid from $39/mo | 24.8M | 1,082 | MCP |
| Pydantic Logfire Observability platform for AI agents, services, and infrastructure with tracing, logging, and metrics. |
10 million spans, logs, or metrics per month | Free plan · paid from $49/mo | 3.9M | 4,519 | |
| Laminar Trace intelligence platform for observing and debugging AI agent runs. |
1 GB data, 7 day access, 1 project, 1 seat | Free plan · paid from $30/mo | 1.1M | — | |
| W&B Weave Tracing, evaluation and monitoring of LLM apps and agents, logged from Python or TypeScript code. |
Free plan: $0/mo, 1 GB/mo Weave data ingestion | Free plan · paid from $60/mo | 439K | 1,142 | MCP |
| Ragas Open-source Python library for evaluating RAG and LLM applications with metrics and test-data generation. |
Open source | not published | 320K | 16.1K | |
| AgentOps Observability and tracing SDK/platform for AI agents: session replay, LLM cost tracking, multi-framework integrations. |
Basic: free up to 5,000 events | Free plan · paid from $40/mo | 19.4K | 5,910 |