Seamlessly connects with your entire AI stack
Splyntra is OpenTelemetry-native and built on open standards. Auto-instrument your favorite agent framework in one line of code, monitor MCP servers, or send OTLP traces from any environment.
Agent Frameworks
Auto-instrument autonomous multi-agent graphs, loops, and tools.
LangGraph
SDK NativeState machine & cyclic graph tracing
CrewAI
SDK NativeHierarchical agent delegation & tasks
OpenAI Agents SDK
SDK NativeAssistant API & multi-tool tracing
LlamaIndex
SDK NativeRAG query engines & tool agents
AutoGen
OTelMulti-agent conversational patterns
Dify / n8n
Webhook / OTelVisual agent & workflow pipelines
Protocols & Open Standards
Open protocols for tool interoperability and zero-lock-in telemetry.
LLM & Foundation Model Providers
Complete token cost calculation, latency percentiles, and cache tracking.
Anthropic Claude
CachedClaude 3.5 Sonnet & 90% prompt cache pricing
OpenAI
ReasoningGPT-4o, o1, o3-mini & reasoning tokens
Google Gemini
2M ContextGemini 1.5 Pro/Flash long context windows
Groq
< 200msUltra-fast inference latency & LPU benchmarks
AWS Bedrock / Azure
VPCEnterprise cloud VPC deployments
Ollama / vLLM
Self-HostedSelf-hosted private models & custom hardware