---
title: "Ingest traces via OTLP"
method: POST
path: "/otlp/v1/traces"
tags: ["OpenTelemetry"]
---

# Ingest traces via OTLP

`POST /otlp/v1/traces`

Ingest traces via the OTLP/HTTP protobuf protocol.

This endpoint accepts a serialized
`ExportTraceServiceRequest` protobuf. Point any OTLP/HTTP
collector or SDK at `POST /otlp/v1/traces` and spans will flow
into the same ingest stream as the Agenta-native endpoints.

Use this when you already have OTel instrumentation emitting
OTLP. For new integrations that don't need raw OTLP, prefer
`POST /tracing/spans/ingest` — it takes JSON, accepts Agenta's
nested shape directly, and surfaces parse failures immediately.

## Content-Type and size limit

Binary protobuf only (`Content-Type: application/x-protobuf`).
JSON OTLP is not accepted. Requests larger than the configured
batch limit (default 4 MB, see `OTLP_MAX_BATCH_BYTES`) return
`413 Request Entity Too Large`.

## Response

Successful ingest returns `200 OK` with a serialized
`ExportTraceServiceResponse` protobuf. Parse failures on the
request body return `400`; malformed spans return `500`; quota
exhaustion returns `403`. Like the native ingest paths, spans
are queued on a Redis stream and persisted asynchronously — see
[Tracing — Async write
contract](/reference/api-guide/tracing#async-write-contract-202).

## Response `200`

Successful Response

- CollectStatusResponse — OTLP endpoint readiness response.
  - `status` string, required — Readiness string. `ready` means the router is mounted and accepts OTLP ingest.

---

[API](https://skmtc.net/agenta-ai/apis/agenta-api.md) · [All operations](https://skmtc.net/agenta-ai/apis/agenta-api/llms.txt) · [OpenAPI document](https://skmtc-service-staging.skmtc.workers.dev/v1/apis/agenta-ai/agenta-api/versions/ea6421d829f1/schema)