SpanlensSpanlens/ Docs
PlatformDocsPricingChangelogBlogGitHub
Getting started
  • Overview
  • Quick start
  • Why Spanlens
Sections
  • Concepts
  • Features
  • Integrations
  • Tutorials
  • Production
  • Migrate to Spanlens
Concepts
  • LLM observability
  • Agent tracing
  • Evals
  • Prompt management
  • Data model
Migrate from
  • Langfuse
  • Helicone
  • LangSmith
Tutorials
  • RAG chatbot
  • Multi-step agent
  • Nightly evals
Core
  • Requests
  • Saved filters
  • Users
  • Traces
Prompts & evaluation
  • Prompts
  • Prompt Playground
  • Prompt A/B
  • Evals
  • Datasets
  • Experiments
  • Annotation
Reliability
  • Security
  • Anomalies
  • Alerts
  • Webhooks
Cost
  • Cost tracking
  • Savings
  • Billing & quotas
Workspace
  • Projects & API keys
  • Shared links
  • Keys & encryption
  • Members & invitations
  • Audit logs
  • Data export
SDK & integrations
  • @spanlens/sdk
  • @spanlens/cli
  • LangChain
  • LangGraph
  • CrewAI
  • OpenAI Assistants
  • AWS Bedrock
  • Instructor
  • Flowise
  • Vercel AI SDK
  • LlamaIndex
  • OpenTelemetry (OTLP)
  • MCP (Cursor / Claude Desktop)
Production
  • Reliability
  • Disaster recovery
  • Scaling
API
  • Direct proxy (any language)
  • REST API reference
  • Error codes
  • Troubleshooting
Self-hosting
  • Docker
  • Backup & restore

Production

Running Spanlens under real traffic: outage behaviour, the fallback queue, recovery runbooks, and tuning for high-throughput workloads.

Reliability

How Spanlens degrades during a partial outage, what the fallback queue does, and how to monitor the proxy so it never silently drops logs.

Disaster recovery

Operator runbook for Spanlens outages: what data is at risk per failure mode, how the fallback queue protects it, and how to recover the Postgres database.

Scaling

Latency budget, log body trade-offs, sampling, and self-hosting tuning for high-throughput LLM workloads on Spanlens.

Back to the docs overview.

SpanlensSpanlens

Observability for people who ship LLM features, not dashboards.

Product
DocsPricingQuick startChangelogBenchmarksBlogStatus
Compare
LangfuseHeliconeLangSmithBraintrustArize PhoenixComet OpikLiteLLMPortkeyAll tools
Guides
LLM ObservabilityAgent TracingLLM Cost TrackingAlternativesCost Calculator
Integrations
OpenAIAnthropicGemini
Open Source
GitHubSelf-hostingSelf-host guide
Company
AboutFAQPrivacyTermsDPASubprocessorsRefund policyAccessibilityContact

© 2026 Spanlens. MIT licensed.

Oceancode · CEO: Haeseong Jeon
Business Reg. No.: 676-71-00622 · Mail-Order Reg.: 2025-Gyeonggi-Gwangju-2133
support@spanlens.io