LangSmith is the observability, debugging, and evaluation platform for LLM applications, built by LangChain. The LangSmith API exposes tracing, dataset management, evaluation, prompt-hub, and Fleet agent functionality for AI engineering teams.
URL: Visit APIs.json URL
Run: Capabilities Using Naftiko
- x-type: company
- AI, LLM, Observability, Evaluations, LangChain
- Created: 2026-05-08
- Modified: 2026-05-08
| API | Description |
|---|---|
| LangSmith Tracing API | Capture and inspect traces for LLM, agent, and chain executions; primary unit of pricing. |
| LangSmith Datasets API | Manage datasets and example records for evaluation. |
| LangSmith Evaluations API | Run offline and online evaluations against datasets, with LLM-as-judge, code-based, and human evaluators. |
| LangSmith Prompt Hub API | Versioned prompt repository for storing, retrieving, and collaborating on LLM prompts. |
| LangSmith Feedback API | Attach human or programmatic feedback to runs and trace nodes. |
| LangSmith Annotation Queues API | Route runs to human reviewers and feed labels back into datasets. |
| LangSmith Fleet (Agent Deployment) API | Deploy and manage LangGraph agents in production via Fleet. |
- Website
- Documentation
- Pricing
- Plans — API Commons Plans 0.1
- RateLimits — API Commons Rate Limits 0.1
- FinOps — FOCUS-aligned FinOps Framework 1.0
| Artifact | Path | Notes |
|---|---|---|
| Plans | plans/langsmith-plans-pricing.yml |
Developer (free) / Plus ($39/seat/mo + usage) / Enterprise |
| Rate Limits | rate-limits/langsmith-rate-limits.yml |
Quota-based: monthly base traces and Fleet runs per plan |
| FinOps | finops/langsmith-finops.yml |
Subscription + Usage billing model |
FN: Kin Lane
Email: kin@apievangelist.com