Langfuse

Langfuse fills the open-source LLM observability and evaluation gap in the catalog. Official sources position it as an LLM engineering platform for tracing agents and applications, prompt management, datasets, experiments, evaluations, cost tracking, OpenTelemetry, SDKs, LiteLLM logging, MCP, CLI, and self-hosting. It should be recommended when readers need transparent traces and eval loops for real AI products rather than only provider dashboards.

• Covers tracing, graphs, sessions, prompts, evals, datasets, experiments, metrics, and cost tracking.
• Open-source and self-hostable, with cloud plans for production teams.
• Integrates with OpenTelemetry, Python/JavaScript SDKs, LiteLLM, MCP, CLI, and coding agents.
• Keep in mind: Teams must understand Langfuse's unit model before comparing it with request- or data-based pricing from competitors.

• Agent building and orchestration: Langfuse applies this capability to LLM observability, Open source workflows so users can move faster while keeping output quality reviewable.
• Model hosting, evaluation, and deployment: Langfuse applies this capability to LLM observability, Open source workflows so users can move faster while keeping output quality reviewable.
• Workflow automation and integrations: Langfuse applies this capability to LLM observability, Open source workflows so users can move faster while keeping output quality reviewable.
• Datasets, demos, and collaboration: Langfuse applies this capability to LLM observability, Open source workflows so users can move faster while keeping output quality reviewable.
• Monitoring, APIs, and production operations: Langfuse applies this capability to LLM observability, Open source workflows so users can move faster while keeping output quality reviewable.

• Open the official website and create a project, workspace, or organization. Keep a human review step in the workflow for facts, privacy, rights, and brand fit.
• Choose a model, agent template, automation flow, or deployment target. Keep a human review step in the workflow for facts, privacy, rights, and brand fit.
• Connect data sources, tools, APIs, and permissions required by the workflow. Keep a human review step in the workflow for facts, privacy, rights, and brand fit.
• Test with realistic inputs, inspect logs, and refine prompts, tools, or policies. Keep a human review step in the workflow for facts, privacy, rights, and brand fit.
• Deploy, monitor, and iterate as usage patterns and reliability requirements evolve. Keep a human review step in the workflow for facts, privacy, rights, and brand fit.

• Langfuse offers a free tier or trial, so you can evaluate it before upgrading.
• Paid plans for Langfuse start at about $29/mo, with higher tiers unlocking more usage, stronger models, and team features.
• Official pricing lists a free Hobby plan with 50k units/month and 30 days of data access, Core at $29/month, Pro at $199/month, and Enterprise at $2,499/month. Additional usage is unit-based.
• Langfuse units count traces, observations, and scores; do not compare the price directly with request-based observability tools without normalizing usage.
• Pricing last checked 2026-07-08, source: https://langfuse.com/pricing. Plans can change, so confirm on the official site.

• Internal workflow automation and operations agents. Langfuse can shorten preparation time, create first drafts, or help teams compare options faster.
• AI application prototyping and model experiments. Langfuse can shorten preparation time, create first drafts, or help teams compare options faster.
• Model hosting, demos, and API-backed products. Langfuse can shorten preparation time, create first drafts, or help teams compare options faster.
• Research pipelines, data processing, and evaluation. Langfuse can shorten preparation time, create first drafts, or help teams compare options faster.
• Customer support, sales operations, and knowledge workflows. Langfuse can shorten preparation time, create first drafts, or help teams compare options faster.

• AI engineers and platform teams. If LLM observability, Open source tasks appear often in your work, Langfuse can become part of a repeatable productivity workflow.
• Automation builders and operations teams. If LLM observability, Open source tasks appear often in your work, Langfuse can become part of a repeatable productivity workflow.
• Startups building AI-native products. If LLM observability, Open source tasks appear often in your work, Langfuse can become part of a repeatable productivity workflow.
• Researchers and model developers. If LLM observability, Open source tasks appear often in your work, Langfuse can become part of a repeatable productivity workflow.
• Enterprises integrating agents into real workflows. If LLM observability, Open source tasks appear often in your work, Langfuse can become part of a repeatable productivity workflow.