
Traceway
An OpenTelemetry-native observability platform combining logs, traces, metrics, exceptions, session replay and AI tracing, self-hosted or in the cloud.
- GitHub stars
- 1.8k
- Last commit
- today
- Latest release
- cli/v2.0.11
- Licence
- MIT
- Self-hosted
- Yes
- Hosted version
- Available

Traceway is an observability platform built natively on OpenTelemetry. It brings logs, traces, metrics, exception tracking, session replay, AI tracing and on-call paging into one product, and you can start by pointing an OTLP exporter at it without a separate collector or a vendor-specific SDK for each language.
Logs are structured and linked to traces, so you can jump from a log line to its span. Traces show end-to-end waterfalls, endpoint views rank routes by latency percentiles and error rate, and exceptions are grouped into ranked issues with source map support. There is experimental profiling with flame graphs, session replay for web and Flutter apps, and LLM observability covering cost, tokens and latency.
The project is written in Go, licensed under MIT and says that every feature ships in the open-source build with no open-core split. You can self-host it for free, using a setup the README says takes about 90 seconds, or run it on Traceway Cloud if you prefer not to manage infrastructure. It suits engineering teams looking for one tool to find and diagnose production problems.
Key features
- Structured logs linked to traces
- End-to-end distributed tracing waterfalls
- Grouped exception tracking with source maps
- Session replay for web and Flutter
- LLM cost and latency observability
- On-call paging and alerting
Pricing: Free to self-host under the MIT licence with all features included; a hosted Traceway Cloud option is also available.
































