telemetry.dev
Trace LLM calls and AI agents with OpenTelemetry.
SubscriptionAICodeMonitoring2 impressionslaunches Oct 211 comment
Comments
>log in to commentWe built Telemetry to give developers a single view of LLM calls, tool steps, and retrievals, so you can see latency, token usage, cost, and errors via OpenTelemetry. It works with our TypeScript/Python SDKs or any OTLP exporter. I’d love your thoughts—please try the "Find the slow step" trace view and let me know what breaks. 😊
More AI tools →74 free dev tools →
Telemetry.dev provides OpenTelemetry-based observability for LLM calls and AI agents, tracking latency, cost, token usage and errors.
- for
- Developers building AI applications and agents who need tracing and cost monitoring.
- pricing
- freemium
Key features
- Unified Traces — Shows model calls, tool steps and retrievals in a single trace with duration per step.
- Cost & Latency Metrics — Displays token usage, spend and latency per model, provider and environment.
- SDKs & Exporters — First-party TypeScript and Python SDKs plus OTLP/HTTP exporter support for any language.
- Prompt Redaction — Option to disable prompt capture per environment to keep sensitive data out of storage.
- Filtering & Search — Search traces by name and filter by status, environment, framework or session.
- Retention Controls — 14-day retention on Free plan, configurable longer retention on paid plans.
Use cases
- Debug slow or failing tool calls in an AI agent.
- Monitor and control LLM spend across multiple providers.
- Compare latency and error rates between model versions.
- Audit token usage and cost per request for compliance.
- Integrate existing OpenTelemetry pipelines to add LLM observability.
telemetry.dev pricing
- Free$0/month30k ingestion units/month · 14-day retention · 1 project · 2 seats
- Pro$25/month500k ingestion units/month · 30-day retention · 5 projects · 5 seats
- Team$50/month2M ingestion units/month · 90-day retention · Unlimited projects · Unlimited seats
telemetry.dev vs alternatives
 telemetry.dev |  SigNoz |  Moda |  Better Stack |  Datadog | |
|---|---|---|---|---|---|
| Best for | LLM tracing with OpenTelemetry | Self-hosted observability | AI agent eval & monitoring | Uptime & incident monitoring | Full-stack cloud monitoring |
| Pricing | Freemium | Free | Subscription | Subscription | Subscription |
| DevHunt upvotes | 1 | 0 | 0 | 0 | 0 |
| Launched | Oct 2031 | — | — | — | — |
- telemetry.dev vs SigNoz: Open-source alternative offering traces, metrics and logs via OpenTelemetry
- telemetry.dev vs Moda: Focuses on production monitoring and evaluation of AI agents rather than generic LLM tracing
- telemetry.dev vs Better Stack: General uptime, logs and incident monitoring, not specialized for LLM traces
- telemetry.dev vs Datadog: Broad cloud-app monitoring platform with paid tiers, not dedicated to LLM observability
telemetry.dev FAQ
What is telemetry.dev?+
It is an LLM observability platform built on OpenTelemetry that groups model calls, tool steps and retrievals into traces with cost, latency and error data.
How do I start sending traces?+
Create a project API key, configure a telemetry.dev integration or an OTLP/HTTP exporter, then run an instrumented request.
Which languages and frameworks are supported?+
First-party SDKs for TypeScript and Python, plus integrations for providers and frameworks like OpenAI, Anthropic, Vercel AI SDK, LangChain, CrewAI, and any language that can send OTLP over HTTP.
Is the free plan really free?+
Yes, it costs $0, requires no credit card, and includes 30,000 ingestion units per month, 14-day retention, one project and two seats.
Can I prevent prompts from being stored?+
Yes, you can disable prompt capture per environment; recognized prompt, response and tool fields are omitted from storage, though they still travel to ingest.
Summarized by DevHunt from telemetry.dev · Oct 28, 2026. Details may change; check the official site.
