> ## Documentation Index
> Fetch the complete documentation index at: https://arizeai-433a7140.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Release Notes

# `get_traces` — Retrieve Traces from a Project

April 1, 2026

**Available in arize-phoenix 13.15.0+ (server), arize-phoenix-client 2.2.0+ (Python)**

`client.traces.get_traces()` retrieves traces for a project with filtering by time range, session, and sort order. The method handles cursor-based pagination automatically, collecting up to `limit` traces across multiple pages.

```python theme={null}
from phoenix.client import Client

client = Client()

# Fetch the 50 most recent traces
traces = client.traces.get_traces(
    project_identifier="my-project",
    limit=50,
)

# Filter by time range and include full span details
from datetime import datetime, timezone

traces = client.traces.get_traces(
    project_identifier="my-project",
    start_time=datetime(2026, 3, 1, tzinfo=timezone.utc),
    end_time=datetime(2026, 4, 1, tzinfo=timezone.utc),
    include_spans=True,
    limit=200,
)

# Filter by session ID
traces = client.traces.get_traces(
    project_identifier="my-project",
    session_id="my-session-id",
)
```

* **`project_identifier`** — project name or ID
* **`start_time` / `end_time`** — inclusive/exclusive bounds on trace start time
* **`sort`** — sort by `"start_time"` (default) or `"latency_ms"`
* **`order`** — `"asc"` or `"desc"` (default `"desc"`)
* **`include_spans`** — when `True`, each trace includes full span detail; use with care on large trace sets
* **`session_id`** — filter to a single session ID or a list of session IDs
* **`limit`** — maximum traces to return; pagination is handled automatically (default `100`)

An async variant is available on `AsyncClient`:

```python theme={null}
from phoenix.client import AsyncClient

client = AsyncClient()
traces = await client.traces.get_traces(
    project_identifier="my-project",
    limit=50,
)
```

# Secrets Management REST API

April 1, 2026

**Available in arize-phoenix 13.21.0+**

Admin users can now store and manage encrypted LLM provider credentials (API keys) in Phoenix via a single REST endpoint. Secrets are encrypted with AES-128-CBC before being persisted and are never returned in API responses.

`PUT /v1/secrets` atomically upserts and deletes secrets in one request. Pass a `value` string to create or update a key; pass `value: null` to delete it.

```json theme={null}
PUT /v1/secrets
{
  "secrets": [
    { "key": "OPENAI_API_KEY", "value": "sk-..." },
    { "key": "ANTHROPIC_API_KEY", "value": "sk-ant-..." },
    { "key": "STALE_KEY", "value": null }
  ]
}
```

Response:

```json theme={null}
{
  "data": {
    "upserted_keys": ["OPENAI_API_KEY", "ANTHROPIC_API_KEY"],
    "deleted_keys": ["STALE_KEY"]
  }
}
```

* **Admin-only** — requires an admin API key or admin session
* **Atomic** — all upserts and deletes in the request succeed or fail together
* **Duplicate keys** — when the same key appears more than once, the last occurrence wins
* **Deleting a non-existent key** succeeds silently

# Python 3.14 Support

April 1, 2026

**Available in arize-phoenix 13.21.0+, arize-phoenix-client 2.2.0+, arize-phoenix-evals 2.13.0+**

Phoenix server, the Python client SDK, and the evals library now support Python 3.14 on Linux and macOS. Windows + Python 3.14 is not yet supported and will raise an explicit error at install time.