> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://www.comet.com/docs/opik/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://www.comet.com/docs/opik/_mcp/server.

# Manage datasets

In Opik 2.0, datasets are project-scoped. Make sure to specify a `project_name` when creating datasets so they are associated with the correct project.

Datasets can be used to track test cases you would like to evaluate your LLM on. Each dataset is made up of a dictionary
with any key value pairs. When getting started, we recommend having an `input` and optional `expected_output` fields for
example. These datasets can be created from:

* Python SDK: You can use the Python SDK to create a dataset and add items to it.
* TypeScript SDK: You can use the TypeScript SDK to create a dataset and add items to it.
* Traces table: You can add existing logged traces (from a production application for example) to a dataset.
* The Opik UI: You can manually create a dataset and add items to it.

Once a dataset has been created, you can run Experiments on it. Each Experiment will evaluate an LLM application based
on the test cases in the dataset using an evaluation metric and report the results back to the dataset.

## Create a dataset via the UI

The simplest and fastest way to create a dataset is directly in the Opik UI.
This is ideal for quickly bootstrapping datasets from CSV files without needing to write any code.

Steps:

1. Navigate to **Evaluation > Datasets** in the Opik UI.
2. Click **Create new dataset**.
3. In the pop-up modal:
   * Provide a name and an optional description
   * Optionally, upload a CSV file with your data
4. Click **Create dataset**.

![](https://fdr-prod-docs-files-public.s3.us-east-1.amazonaws.com/opik.docs.buildwithfern.com/4676fe7d51059dc510b015c4e6a494a39fceb7394d15173632faea95de7bff76/img/evaluation/create_dataset.png?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Content-Sha256=UNSIGNED-PAYLOAD&X-Amz-Credential=AKIA6KXJSKKNFOCF7G4B%2F20260922%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20260922T061223Z&X-Amz-Expires=604800&X-Amz-Signature=b2ce1dac3ec8a6c7a4aab259a0687b3f2ef6ec0ba8aac4dfe6c3ff44efe87585&X-Amz-SignedHeaders=host&x-amz-checksum-mode=ENABLED&x-id=GetObject)

If you need to create a dataset with more than 1,000 rows, you [can use the SDK](/evaluation/advanced/manage_datasets#creating-a-dataset-using-the-sdk).

The UI dataset creation has some limitations:

* File size is limited to 1,000 rows via the UI.
* No support for nested JSON structures in the CSV itself.

For datasets requiring rich metadata, complex schemas, or programmatic control, use the SDK instead (see the next section).

When you create a dataset with a CSV file, this creates the first version (v1)
of your dataset. All subsequent modifications will create new versions automatically.

## Understanding dataset versioning

Dataset versioning in Opik creates **immutable snapshots** of your data. Every time you modify a dataset—whether adding, editing, or deleting items—a new version is automatically created. This ensures complete reproducibility, provides an audit trail of all changes, and allows easy rollback to any previous state.

Each dataset version contains:

* **Version name**: Auto-generated sequential name (v1, v2, v3, etc.)
* **Change description**: Optional note describing what changed
* **Tags**: Labels for categorizing versions (e.g., `production`, `baseline`)
* **Item statistics**: Count of items added, modified, and deleted
* **Timestamp and author**: When the version was created and by whom

Once a version is created, its data cannot be changed—any modification creates a new version instead. Restoring a previous version also creates a *new* version with the same data, preserving your complete version timeline.

The special `latest` tag always points to the most recent version.
When running experiments without specifying a version, `latest` is used by default.

## Working with draft mode (UI)

When making changes to a dataset in the Opik UI, all modifications go into a **draft state** first. This gives you a staging area to review changes before committing them as a new version. The draft is visible only to you, and AI-generated samples from "Expand with AI" also go to draft for review.

When a dataset has unsaved draft changes, an orange **"Draft"** tag appears next to the dataset name, and **Save changes** / **Discard changes** buttons appear in the toolbar. Items show colored borders: green for newly added items, amber for modified items.

![](https://fdr-prod-docs-files-public.s3.us-east-1.amazonaws.com/opik.docs.buildwithfern.com/27a8ac8ccf3ed750a6488efefc2c04a928713b213f1b2e1ec5124847ca7968c4/img/evaluation/dataset_draft_mode.png?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Content-Sha256=UNSIGNED-PAYLOAD&X-Amz-Credential=AKIA6KXJSKKNFOCF7G4B%2F20260922%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20260922T061223Z&X-Amz-Expires=604800&X-Amz-Signature=a96f65aafda55b7e611e5c7cd61bcfa62d2a43cbf7c02e955c89b03fc1cf0335&X-Amz-SignedHeaders=host&x-amz-checksum-mode=ENABLED&x-id=GetObject)

### Saving or discarding changes

To commit your draft as a new version:

1. Click **Save changes** in the toolbar
2. Enter a **version note** describing what changed
3. Optionally add **tags** to categorize this version
4. Click **Save**

![](https://fdr-prod-docs-files-public.s3.us-east-1.amazonaws.com/opik.docs.buildwithfern.com/28cc5a161d0fe6b2d652d2fedb4e323d00102275fd570b9e54aa1b58d458e7b7/img/evaluation/save_version_dialog.png?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Content-Sha256=UNSIGNED-PAYLOAD&X-Amz-Credential=AKIA6KXJSKKNFOCF7G4B%2F20260922%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20260922T061223Z&X-Amz-Expires=604800&X-Amz-Signature=4995b6174b3fd2f17882860b7c822bd4c804eea30d1fe8c9bd4a74d7ea5624e7&X-Amz-SignedHeaders=host&x-amz-checksum-mode=ENABLED&x-id=GetObject)

To abandon your draft, click **Discard changes** and confirm. If you try to navigate away with unsaved changes, Opik displays a warning to prevent accidental loss of work.

Use draft mode to batch related changes into a single, well-documented version.

## Version history

To view the complete timeline of dataset changes, navigate to your dataset and click the **Version history** tab. The table shows each version's name, change summary (items added/modified/deleted), version note, tags, item count, and creation timestamp.

![](https://fdr-prod-docs-files-public.s3.us-east-1.amazonaws.com/opik.docs.buildwithfern.com/52f8d0087c5341889adb67be721e07c57604f0cf9adaa67b4941e8d7a709a95b/img/evaluation/version_history_tab.png?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Content-Sha256=UNSIGNED-PAYLOAD&X-Amz-Credential=AKIA6KXJSKKNFOCF7G4B%2F20260922%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20260922T061223Z&X-Amz-Expires=604800&X-Amz-Signature=bc6166fb932db5adf71c75b579c66c487468e2e2847dd14cc33dfed4c8271314&X-Amz-SignedHeaders=host&x-amz-checksum-mode=ENABLED&x-id=GetObject)

From this view you can:

* **View items**: Click a version row and select **View items** to see the exact data at that point in time
* **Restore**: Click the **⋮** menu and select **Restore this version** to create a new version with that data
* **Edit metadata**: Click the **⋮** menu and select **Edit** to update the version note or tags (the data itself remains immutable)

Restoring a version creates a **new** version with the same data.
No history is lost or overwritten.

### Managing dataset and version tags from the SDK

The `Dataset` object exposes `get_tags()` to read the current tags, but does not yet provide a dedicated setter. To write tags programmatically — for example to drive an `env:prod` / `env:stage` promotion workflow — use the REST client exposed on the Opik client.

There are two tag surfaces, depending on what you want to scope the tag to:

* **Dataset-level tags** apply to the dataset as a whole and persist across versions. Use `update_dataset` — this **replaces** the existing tag list.
* **Version-level tags** apply to a specific dataset version. Use `update_dataset_version` — this is **additive** (it adds to the version's existing tags).

```python {pytest_codeblocks_skip=true}
import opik

client = opik.Opik()
dataset = client.get_or_create_dataset(name="my-eval", project_name="my-project")

# Read current dataset-level tags
print(dataset.get_tags())

# Set dataset-level tags (replaces the existing list)
client.rest_client.datasets.update_dataset(
    id=dataset.id,
    name=dataset.name,
    tags=["env:prod"],
)

# Add tags to a specific version (additive)
client.rest_client.datasets.update_dataset_version(
    id=dataset.id,
    version_hash=dataset.version_hash,
    tags_to_add=["env:prod"],
)
```

`client.rest_client` is a thin wrapper around the public REST API. The underlying endpoints are stable, but the Python wrapper itself is not guaranteed to be backward-compatible across SDK versions. A first-class `Dataset.set_tags()` / `Dataset.add_version_tags()` helper is on the roadmap — this snippet is the supported interim path.

You can then filter dataset items by these tags via [`get_items(filter_string=...)`](#querying-dataset-items) using the `tags contains` operator.

## Adding traces to a dataset

One of the most powerful ways to build evaluation datasets is by converting production traces into dataset items. This allows you to leverage real-world interactions from your LLM application to create test cases for evaluation.

### Adding traces via the UI

To add traces to a dataset from the Opik UI:

1. Navigate to the traces page
2. Select one or more traces you want to add to a dataset
3. Click the **Add to dataset** button in the toolbar
4. In the dialog that appears:
   * Select an existing dataset or create a new one
   * Choose which trace metadata to include:
     * **Nested spans**: Include all child spans within the trace
     * **Tags**: Include trace tags
     * **Feedback scores**: Include any feedback scores attached to the trace
     * **Comments**: Include comments added to the trace
     * **Usage metrics**: Include token usage and cost information
     * **Metadata**: Include custom metadata fields
5. Click on the dataset name to add the selected traces

![Add traces to dataset modal](https://fdr-prod-docs-files-public.s3.us-east-1.amazonaws.com/opik.docs.buildwithfern.com/3dde53f377293d0aeccbe5d5fa6ef19c4da929c0eba5c2e7b1822e2d21d56a6e/img/evaluation/add_traces_to_dataset.png?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Content-Sha256=UNSIGNED-PAYLOAD&X-Amz-Credential=AKIA6KXJSKKNFOCF7G4B%2F20260922%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20260922T061223Z&X-Amz-Expires=604800&X-Amz-Signature=a75d1bfe8b22c63f624acd1ac1e3df46929fb8530bb43b62c96a12783fdaedd3&X-Amz-SignedHeaders=host&x-amz-checksum-mode=ENABLED&x-id=GetObject)

By default, all metadata options are enabled. You can uncheck any options you don't need. The trace's input and output are always included.

### What gets added to the dataset

When you add a trace to a dataset, the following structure is created:

* **input**: The trace's input data
* **expected\_output**: The trace's output data (stored as `expected_output` for evaluation purposes)
* **spans** (optional): Array of nested spans with their inputs, outputs, and metadata
* **tags** (optional): Array of tags associated with the trace
* **feedback\_scores** (optional): Array of feedback scores with name, value, and source
* **comments** (optional): Array of comments with text and ID
* **usage** (optional): Token usage and cost information
* **metadata** (optional): Custom metadata fields

This rich structure allows you to:

* Evaluate complex multi-step workflows by including nested spans
* Filter and analyze based on tags and metadata
* Use existing feedback scores as ground truth for evaluation
* Preserve context through comments and annotations

## Creating a dataset using the SDK

In Opik 2.0, datasets are project-scoped. Specify a `project_name` to associate your dataset with the correct project.

You can create a dataset and log items to it using the `get_or_create_dataset` method:

**`TypeScript SDK`**

```typescript title="TypeScript SDK" language="typescript"
import { Opik } from "opik";

// Create a dataset
const client = new Opik();
const dataset = await client.getOrCreateDataset("My dataset", "Evaluation dataset", "my-project");
```

**`Python SDK`**

```python title="Python SDK" language="python"
from opik import Opik

# Create a dataset
client = Opik()
dataset = client.get_or_create_dataset(name="My dataset", project_name="my-project")
```

If a dataset with the given name already exists, the existing dataset will be returned.

### Insert items

#### Inserting dictionary items

You can insert items to a dataset using the `insert` method:

**`TypeScript`**

```typescript title="TypeScript" language="typescript"
import { Opik } from "opik";
const client = new Opik();
const dataset = await client.getOrCreateDataset("My dataset", "Evaluation dataset", "my-project");

dataset.insert([
  { user_question: "Hello, world!", expected_output: { assistant_answer: "Hello, world!" } },
  { user_question: "What is the capital of France?", expected_output: { assistant_answer: "Paris" } },
]);
```

**`Python`**

```python title="Python" language="python"
import opik

# Get or create a dataset
client = opik.Opik()
dataset = client.get_or_create_dataset(name="My dataset", project_name="my-project")

# Add dataset items to it
dataset.insert([
    {"user_question": "Hello, world!", "expected_output": {"assistant_answer": "Hello, world!"}},
    {"user_question": "What is the capital of France?", "expected_output": {"assistant_answer": "Paris"}},
])
```

Opik automatically deduplicates items that are inserted into a dataset when using the Python SDK. This means that you
can insert the same item multiple times without duplicating it in the dataset. This combined with the `get or create
  dataset` methods means that you can use the SDK to manage your datasets in a "fire and forget" manner. It can be
turned off with `deduplication=False`, see [Disabling deduplication](#disabling-deduplication).

#### Disabling deduplication

Deduplication requires the Python SDK to download the dataset's existing items once so it can compare their
content hashes against the items you are inserting. On large datasets that download dominates the insert. If you
already know your items are unique — for example when populating a fresh dataset, or when you generate ids
yourself — pass `deduplication=False` to skip that work entirely: nothing is downloaded, no hashes are computed,
and every item you pass is sent as-is.

**`Python`**

```python title="Python" language="python"
dataset.insert(items, deduplication=False)
```

The flag is available on every Python SDK method that writes items — `insert`, `update`, `insert_from_json`,
`insert_from_pandas` and `read_jsonl_from_file` — as well as on the equivalent `TestSuite` methods. With
deduplication disabled, inserting the same content twice produces two separate dataset items.

When using the SDK to insert items, a new dataset version is automatically created.
If you insert items in multiple batches within a single `insert()` call, they are grouped into one version.
The Python SDK uploads those batches on 8 worker threads by default; use `num_threads=1` to upload them
sequentially instead. See [Tuning SDK throughput](#tuning-sdk-throughput). Parallel upload requires a recent
Opik backend — against older ones the SDK falls back to a sequential upload and logs a warning.

Once the items have been inserted, you can view them in the Opik UI:

![](https://fdr-prod-docs-files-public.s3.us-east-1.amazonaws.com/opik.docs.buildwithfern.com/5533ca8603ece41e5a62ef6a89e1b4e9eb788a46ca5f650ab4d76d8076272845/img/evaluation/dataset_items_page.png?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Content-Sha256=UNSIGNED-PAYLOAD&X-Amz-Credential=AKIA6KXJSKKNFOCF7G4B%2F20260922%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20260922T061223Z&X-Amz-Expires=604800&X-Amz-Signature=4e07fb7349a0fae34652e1d69d72dd13529f2b05a3906a658b4c9e57f8147131&X-Amz-SignedHeaders=host&x-amz-checksum-mode=ENABLED&x-id=GetObject)

#### Inserting items from a JSONL file

You can also insert items from a JSONL file:

**`Python`**

```python title="Python" language="python"
import opik

client = opik.Opik()
dataset = client.get_or_create_dataset(name="My dataset", project_name="my-project")

dataset.read_jsonl_from_file("path/to/file.jsonl")

```

#### Inserting items from a Pandas DataFrame

You can also insert items from a Pandas DataFrame:

**`Python`**

```python title="Python" language="python"
import opik

client = opik.Opik()
dataset = client.get_or_create_dataset(name="My dataset", project_name="my-project")

dataset.insert_from_pandas(dataframe=df)

# You can also specify an optional keys_mapping parameter
dataset.insert_from_pandas(dataframe=df, keys_mapping={"Expected output": "expected_output"})
```

### Deleting items

You can delete items in a dataset by using the `delete` method:

**`TypeScript`**

```typescript title="TypeScript" language="typescript"
import { Opik } from "opik";

// Get or create a dataset
client = new Opik();
dataset = await client.getDataset("My dataset")

await dataset.delete(["123", "456"])

// Or to delete all items
await dataset.clear()
```

**`Python`**

```python title="Python" language="python"
from opik import Opik

# Get or create a dataset
client = Opik()
dataset = client.get_dataset(name="My dataset")

dataset.delete(items_ids=["123", "456"])

# Or to delete all items
dataset.clear()
```

Deleting items creates a new version of the dataset. The deleted items remain accessible
in previous versions through the version history, ensuring you never permanently lose data.

## Downloading a dataset from Opik

You can download a dataset from Opik using the `get_dataset` method:

**`TypeScript`**

```typescript title="TypeScript" language="typescript"
import { Opik } from "opik";

const client = new Opik();
const dataset = await client.getDataset("My dataset");

const items = await dataset.getItems();
console.log(items);
```

**`Python`**

```python title="Python" language="python"
from opik import Opik

client = Opik()
dataset = client.get_dataset(name="My dataset")

# Get items as list of DatasetItem objects
items = dataset.get_items()

# Convert to a Pandas DataFrame
dataset.to_pandas()

# Convert to a JSON array
dataset.to_json()
```

### Downloading large datasets faster

Dataset items are fetched a page at a time, and those pages are downloaded concurrently. The default is
8 threads; pass `num_threads` to read with more or fewer (see [Tuning SDK throughput](#tuning-sdk-throughput)):

**`Python`**

```python title="Python" language="python"
from opik import Opik

client = Opik()
dataset = client.get_dataset(name="My dataset")

# The whole dataset as one list, downloaded over 16 threads
items = dataset.get_items(num_threads=16)
```

The thread count never changes the result, only how quickly it arrives.

`get_items()` returns the whole dataset as a single list, so the call does not return until every
item has been downloaded and the full result is held in memory. `stream_items()` reads the same
items in chunks instead, yielding each chunk as soon as it arrives. Use it when you want to start
processing before the download finishes, or when the dataset is too large to hold in memory all at
once:

**`Python`**

```python title="Python" language="python"
from opik import Opik

client = Opik()
dataset = client.get_dataset(name="My dataset")

# One chunk at a time, instead of the whole dataset at once
for chunk in dataset.stream_items(chunk_size=5000):
    process(chunk)  # a list of dicts, exactly as get_items() returns them
```

Chunks arrive in dataset order; only the last one may be shorter than `chunk_size`. Both methods
accept the same `filter_string`, and `nb_samples` to read only the first N items:

**`Python`**

```python title="Python" language="python"
for chunk in dataset.stream_items(
    filter_string='data.category = "geography"',
    nb_samples=10_000,
):
    process(chunk)
```

`nb_samples` must be a positive integer — omit it or pass `None` to read everything. Passing `0`
or a negative value raises `ValueError` rather than being treated as a limit.

`chunk_size` controls how many items each request fetches. It defaults to 2000, which is also
the maximum — a larger value raises `ValueError`, so that peak memory stays bounded. Fetching a
chunk costs a fixed overhead whatever its size, so lowering it makes the whole read slower;
lower it only when individual items are large, since up to `2 * num_threads` chunks are held in
memory at once.

## Tuning SDK throughput

The Python SDK moves dataset and experiment items on a pool of worker threads. Every one of these
paths defaults to **8 threads**, which is what we benchmark against and what we recommend for
customer-scale datasets — you should not need to pass `num_threads` at all:

| Operation        | Call                                            | Default |
| ---------------- | ----------------------------------------------- | ------- |
| Dataset read     | `dataset.get_items()`, `dataset.stream_items()` | 8       |
| Dataset write    | `dataset.insert()`                              | 8       |
| Experiment write | `experiment.batch_upload_items()`               | 8       |

Each of those takes a `num_threads` argument if you do want to change it:

**`Python`**

```python title="Python" language="python"
items = dataset.get_items(num_threads=4)       # read more gently
dataset.insert(items, num_threads=16)          # more concurrency, if the client has CPU headroom
experiment.batch_upload_items(records, num_threads=1)  # upload sequentially
```

**Raise it** when the client is idle waiting on the network — a big upload over a high-latency link
is the usual case. Do not expect much: on a 119,903-item upload we measured 16 threads running
slightly *slower* than 8, because the SDK saturates a CPU core serializing and compressing payloads
long before thread count becomes the limit. Past 8, extra workers mostly add scheduling overhead.

**Lower it** when you are sharing a rate limit with other jobs, when the client machine is small, or
when individual items are large enough that memory matters. On the dataset paths, `num_threads`
bounds memory as well as concurrency: a read holds up to `2 * num_threads` chunks and an upload up
to `2 * num_threads` request bodies, whatever the dataset's size. `experiment.batch_upload_items()`
does not work that way — it builds every batch up front and queues them all, so its peak memory
tracks the total number of items and lowering `num_threads` will not contain it. Split the records
across calls if an experiment upload is too large to hold.

`num_threads=1` makes the operation fully sequential, which is the only setting that guarantees
batches arrive in order.

`num_threads` must be a positive integer. Dataset reads and experiment uploads cap it at 32, since
the SDK's HTTP client pools a limited number of connections and extra threads would queue behind it
rather than speed anything up. Thread count never changes the result of an operation that succeeds,
only how fast it arrives. It does change what a *failure* leaves behind — see below.

`update()`, `insert_from_json()`, `insert_from_pandas()` and `read_jsonl_from_file()` all upload
through `insert()`, so they get the same 8-thread default — but they do not take a `num_threads`
argument of their own, and passing one raises `TypeError`. Convert to items and call `insert()`
directly when you need to tune one of those uploads.

On a failed upload, the worker count decides what was already written, and the two write paths
differ. With `num_threads=1` both stop cleanly: nothing after the failed batch is sent.

In parallel, `dataset.insert()` still **sends everything it has already queued** — the call drains
the in-flight work and waits for it before re-raising, so the exception surfaces after those
batches have landed, not before. `experiment.batch_upload_items()` does the opposite: batches that
have not started are dropped and the ones already running are not waited for, so a few batches
after the failed one may or may not have landed.

Neither path rolls back what already succeeded. Use `num_threads=1` when you need a failure to
stop at a predictable point.

Parallel dataset upload also requires an Opik backend of 2.2.8 or newer. Against an older backend
the SDK falls back to a sequential upload and logs a warning, whatever `num_threads` you pass.

## Filtering datasets programmatically

You can filter dataset items using the `filter_string` parameter on the `get_items()` method or when
running evaluations with `evaluate_prompt()`. This allows you to work with specific subsets of your data.

### Basic filtering

**`Python`**

```python title="Python" language="python"
from opik import Opik

client = Opik()
dataset = client.get_dataset(name="my_dataset")

# Get filtered items
failed_items = dataset.get_items(filter_string='tags contains "failed"')

```

### Filter syntax

The filter string uses Opik Query Language (OQL) syntax. Supported columns include:

| Column            | Type       | Description                                                    |
| ----------------- | ---------- | -------------------------------------------------------------- |
| `id`              | String     | Unique identifier for the dataset item                         |
| `source`          | String     | Source of the dataset item                                     |
| `trace_id`        | String     | Associated trace ID                                            |
| `span_id`         | String     | Associated span ID                                             |
| `data`            | Dictionary | Use dot notation for nested fields (e.g., `data.category`)     |
| `tags`            | List       | Use "contains" operator (e.g., `tags contains "test"`)         |
| `created_at`      | DateTime   | ISO 8601 format (e.g., `created_at >= "2024-01-01T00:00:00Z"`) |
| `last_updated_at` | DateTime   | ISO 8601 format                                                |
| `created_by`      | String     | User who created the item                                      |
| `last_updated_by` | String     | User who last updated the item                                 |

### Filter examples

**`Python`**

```python title="Python" language="python"
from opik import Opik

client = Opik()
dataset = client.get_dataset(name="my_dataset")

# Filter by tag
failed_items = dataset.get_items(filter_string='tags contains "failed"')

# Filter by data field
finance_items = dataset.get_items(filter_string='data.category = "finance"')

# Filter by date
recent_items = dataset.get_items(
    filter_string='created_at >= "2024-06-01T00:00:00Z"'
)

# Multiple conditions
filtered_items = dataset.get_items(
    filter_string='tags contains "production" AND data.difficulty = "hard"'
)
```

## Running experiments with dataset versions

When you run an experiment, Opik automatically links it to the specific dataset version that was used. This ensures complete reproducibility—you can always know exactly which data was used for any experiment.

### Automatic version association

Every experiment records which dataset version it used:

* When running from the UI or SDK without specifying a version, the `latest` version is used
* The experiment results page shows the associated dataset version
* You can click the version to see the exact data that was evaluated

This association is permanent. Even if you later modify the dataset, your experiment results remain linked to the original version used.

### Selecting a specific version in Playground

When running experiments from the Playground:

1. Open the Playground and configure your prompt
2. In the dataset selector, choose your dataset
3. A nested dropdown appears showing available versions
4. Select the specific version you want to use, or choose `latest` for the most recent

![](https://fdr-prod-docs-files-public.s3.us-east-1.amazonaws.com/opik.docs.buildwithfern.com/cfcb780469cce1f7999db80b0cd4b8467af4ef3585d2eafa49f0150629c57b4d/img/evaluation/dataset_version_select.png?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Content-Sha256=UNSIGNED-PAYLOAD&X-Amz-Credential=AKIA6KXJSKKNFOCF7G4B%2F20260922%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20260922T061223Z&X-Amz-Expires=604800&X-Amz-Signature=9b7521571af514cb651d34aa0bd59897b0d9a44e95332ffa5a4b852101e4cf90&X-Amz-SignedHeaders=host&x-amz-checksum-mode=ENABLED&x-id=GetObject)

When comparing experiments or running A/B tests, use the same dataset version
to isolate the effect of your changes. This ensures differences in results
are due to your prompt or model changes, not data variations.

### Selecting a specific version in the SDK

When running experiments programmatically, you can specify which dataset version to use by passing a `DatasetVersion` object to `evaluate()`:

**`Python`**

```python title="Python" language="python"
from opik import Opik
from opik.evaluation import evaluate

client = Opik()
dataset = client.get_dataset(name="My dataset")

# Run experiment on the latest version (default behavior)
result = evaluate(
    experiment_name="baseline-experiment",
    dataset=dataset,
    task=my_task_function,
    scoring_metrics=[my_metric],
    project_name="my-project",
)

# Run experiment on a specific version
v1_view = dataset.get_version_view("v1")
result = evaluate(
    experiment_name="v1-experiment",
    dataset=v1_view,  # Pass the DatasetVersion object
    task=my_task_function,
    scoring_metrics=[my_metric],
    project_name="my-project",
)
```

**`TypeScript`**

```typescript title="TypeScript" language="typescript"
import { Opik, evaluate } from "opik";

const client = new Opik();
const dataset = await client.getDataset("My dataset");

// Run experiment on the latest version (default)
const result = await evaluate({
  experimentName: "baseline-experiment",
  dataset: dataset,
  task: myTaskFunction,
  scoringMetrics: [myMetric],
  projectName: "my-project",
});

// Run experiment on a specific version
const v2 = await dataset.getVersionView("v2");
const pinnedResult = await evaluate({
  experimentName: "pinned-experiment",
  dataset: v2,
  task: myTaskFunction,
  scoringMetrics: [myMetric],
  projectName: "my-project",
});
```

### Working with dataset versions programmatically

The SDK provides methods for inspecting and working with dataset versions:

**`Python`**

```python title="Python" language="python"
from opik import Opik

client = Opik()
dataset = client.get_dataset(name="My dataset")

# Get the current (latest) version name
current_version = dataset.get_current_version_name()
print(f"Current version: {current_version}")  # e.g., "v3"

# Get detailed version info (returns DatasetVersionPublic)
version_info = dataset.get_version_info()
print(f"Version ID: {version_info.id}")
print(f"Version name: {version_info.version_name}")
print(f"Items total: {version_info.items_total}")
print(f"Created at: {version_info.created_at}")

# Get a read-only view of a specific version
v1_view = dataset.get_version_view("v1")

# Access version metadata
print(f"Version: {v1_view.version_name}")
print(f"Items in v1: {v1_view.items_total}")
print(f"Items added: {v1_view.items_added}")
print(f"Items modified: {v1_view.items_modified}")
print(f"Items deleted: {v1_view.items_deleted}")

# Get items from a specific version
v1_items = v1_view.get_items()

# Export version data
v1_df = v1_view.to_pandas()
v1_json = v1_view.to_json()
```

**`TypeScript`**

```typescript title="TypeScript" language="typescript"
import { Opik } from "opik";

const client = new Opik();
const dataset = await client.getDataset("My dataset");

// Get the current (latest) version name
const currentVersion = await dataset.getCurrentVersionName();
console.log(`Current version: ${currentVersion}`); // e.g., "v3"

// Get detailed version info (returns DatasetVersionPublic)
const versionInfo = await dataset.getVersionInfo();
console.log(`Version ID: ${versionInfo?.id}`);
console.log(`Version name: ${versionInfo?.versionName}`);
console.log(`Items total: ${versionInfo?.itemsTotal}`);
console.log(`Created at: ${versionInfo?.createdAt}`);

// Get a read-only view of a specific version
const v1View = await dataset.getVersionView("v1");

// Access version metadata
console.log(`Version: ${v1View.versionName}`);
console.log(`Items in v1: ${v1View.itemsTotal}`);
console.log(`Items added: ${v1View.itemsAdded}`);
console.log(`Items modified: ${v1View.itemsModified}`);
console.log(`Items deleted: ${v1View.itemsDeleted}`);

// Get items from a specific version
const v1Items = await v1View.getItems();

// Export version data as JSON
const v1Json = await v1View.toJson();
```

`DatasetVersion` is a read-only view. You cannot insert, update, or delete items
through a `DatasetVersion` object. All mutations must be done through the `Dataset` object.

## Expanding a dataset with AI

Dataset expansion allows you to use AI to generate additional synthetic samples based on your existing dataset. This is particularly useful when you have a small dataset and want to create more diverse test cases to improve your evaluation coverage.

The AI analyzes the patterns in your existing data and generates new samples that follow similar structures while introducing variations. This helps you:

* **Increase dataset size** for more comprehensive evaluation
* **Create edge cases** and variations you might not have considered
* **Improve model robustness** by testing against diverse inputs
* **Scale your evaluation** without manual data creation

### How to expand a dataset

To expand a dataset with AI:

1. **Navigate to your dataset** in the Opik UI (Evaluation > Datasets > \[Your Dataset])
2. **Click the "Expand with AI" button** in the dataset view
3. **Configure the expansion settings**:
   * **Model**: Choose the LLM model to use for generation (supports GPT-4, GPT-5, Claude, and other models)
   * **Sample Count**: Specify how many new samples to generate (1-100)
   * **Preserve Fields**: Select which fields from your original data to keep unchanged
   * **Variation Instructions**: Provide specific guidance on how to vary the data (e.g., "Create variations that test edge cases" or "Generate examples with different complexity levels")
   * **Custom Prompt**: Optionally provide a custom prompt template instead of the auto-generated one
4. **Start the expansion** - The AI will analyze your data and generate new samples
5. **Review the results** - Generated samples are added to your **draft**. You can review, edit, or remove them before saving to create a new version

![](https://fdr-prod-docs-files-public.s3.us-east-1.amazonaws.com/opik.docs.buildwithfern.com/2d6d780ffc17285e39bc00c0797a1fc2cbbdf851cee270b7f569ebe5c2e9bc1a/img/evaluation/dataset_expansion_modal.png?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Content-Sha256=UNSIGNED-PAYLOAD&X-Amz-Credential=AKIA6KXJSKKNFOCF7G4B%2F20260922%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20260922T061223Z&X-Amz-Expires=604800&X-Amz-Signature=61b3d58a1d5f773415d2ee4392ecf488e46f3aaf2927e98152657298401daf5d&X-Amz-SignedHeaders=host&x-amz-checksum-mode=ENABLED&x-id=GetObject)

### Configuration options

**Sample Count**: Start with a smaller number (10-20) to review the quality before generating larger batches.

**Preserve Fields**: Use this to maintain consistency in certain fields while allowing variation in others. For example, preserve the `category` field while varying the `input` and `expected_output`.

**Variation Instructions**: Provide specific guidance such as:

* "Create variations with different difficulty levels"
* "Generate edge cases and error scenarios"
* "Add examples with different input formats"
* "Include multilingual variations"

### Best practices

* **Start small**: Generate 10-20 samples first to evaluate quality before scaling up
* **Review generated content**: Always review AI-generated samples for accuracy and relevance
* **Use variation instructions**: Provide clear guidance on the type of variations you want
* **Preserve key fields**: Use field preservation to maintain important categorizations or metadata
* **Iterate and refine**: Use the custom prompt option to fine-tune generation for your specific needs

Dataset expansion works best when you have at least 5-10 high-quality examples in your original dataset. The AI uses
these examples to understand the patterns and generate similar but varied content.

## Managing dataset item tags

Tags are a powerful way to organize, categorize, and filter your dataset items. You can use tags to:

* **Categorize test cases** by type, difficulty, or domain (e.g., `edge-case`, `production`, `multilingual`)
* **Track data sources** where items originated from (e.g., `user-feedback`, `synthetic`, `real-world`)
* **Mark review status** during dataset curation (e.g., `needs-review`, `validated`, `archived`)
* **Filter for evaluation** to run experiments on specific subsets of your data
* **Organize workflows** by marking items for different stages or teams

Each dataset item can have multiple tags.

### Adding tags to dataset items

#### Adding tags to individual items

To add tags to a single dataset item:

1. **Navigate to your dataset** in the Opik UI (Evaluation > Datasets > \[Your Dataset])
2. **Click on any dataset item** to open the details panel
3. **In the Tags section**, click the **"+" button**
4. **Type the tag name** and press Enter
5. The tag will be immediately added and saved

You can remove tags by clicking the **"×" icon** next to any tag in the details panel.

#### Adding tags to multiple items (batch operation)

To add the same tag to multiple dataset items at once:

1. **Navigate to your dataset** in the Opik UI
2. **Select multiple items** by clicking the checkboxes next to each item
3. **Click the "Add tags" button** in the toolbar (visible when items are selected)
4. **Enter the tag name** in the dialog that appears
5. **Click "Add tag"** to apply the tag to all selected items

This is particularly useful when you want to categorize a group of related test cases or mark items from the same data source.

Tags are case-sensitive and support alphanumeric characters, hyphens, and underscores. Choose consistent naming conventions for your tags to make filtering easier.

### Filtering dataset items by tags

Once you've tagged your dataset items, you can filter them to work with specific subsets:

1. **Navigate to your dataset** in the Opik UI
2. **Click the "Filters" button** next to the search bar
3. **Select "Tags" from the Column dropdown**
4. **Choose "contains" as the operator**
5. **Enter the tag name** you want to filter by
6. **Close the dialog** to apply the filter

The dataset items table will update to show only items matching your filter criteria. You can:

* **View filtered items** to focus on specific categories
* **Run experiments** on filtered subsets by using the filtered view
* **Export filtered data** for specific test case groups
* **Combine with other filters** to create complex queries

The filter is saved in the URL, so you can bookmark or share specific filtered views of your dataset.

## Bulk operations

Opik supports bulk operations for efficiently managing large datasets. These operations help you work with many items at once without tedious individual selections.

### Select all functionality

When working with datasets that span multiple pages:

1. **Select items on the current page** using the checkbox in the table header
2. A banner appears offering to **"Select all items"** across all pages
3. Click to select all items matching your current filter criteria

![](https://fdr-prod-docs-files-public.s3.us-east-1.amazonaws.com/opik.docs.buildwithfern.com/c42b2ffedb2797cbfcdbee35bb7116da0fb3e8614fe4cca66e1e10093537b47a/img/evaluation/dataset_bulk_select.png?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Content-Sha256=UNSIGNED-PAYLOAD&X-Amz-Credential=AKIA6KXJSKKNFOCF7G4B%2F20260922%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20260922T061223Z&X-Amz-Expires=604800&X-Amz-Signature=c6c9330dde41bd9f546d42de099708030b70aa916c0ec417f3fb5df0e6ccb629&X-Amz-SignedHeaders=host&x-amz-checksum-mode=ENABLED&x-id=GetObject)

This works with filtered views too—if you have a filter applied, "Select all" only selects items matching that filter.

### Available bulk operations

Once you have items selected, the toolbar shows available operations:

* **Add tags**: Apply one or more tags to all selected items
* **Delete**: Remove selected items (creates a new version with items removed)
* **Export**: Download selected items as CSV or JSON

### Processing indicators

For large bulk operations:

* A loading indicator shows "Your dataset is still processing..."
* The operation runs in the background—you can continue browsing
* A success message appears when processing completes

For very large datasets, bulk operations are processed in batches. The UI remains
responsive during processing, and you'll see progress indicators for long-running operations.

**Recommended if you build with an AI coding assistant.** Assembling a dataset by hand is the
slowest part of setting up an evaluation. One command — `opik configure` — installs both the
[MCP server](/mcp-server) and the Opik skills, and your assistant can then create the dataset and
fill it in for you — writing the first cases from scratch, or, if you are already logging traces,
pulling them from the ones that scored badly.

An example prompt:

*"Build an Opik dataset from the traces scored below 0.7 on answer relevance, fill in the expected
answers, then evaluate my agent against it."*