> ## Documentation Index
> Fetch the complete documentation index at: https://exa.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

> ## Agent Instructions
> The Exa API is served at https://api.exa.ai. Authenticate with `Authorization: Bearer $EXA_API_KEY` (or `x-api-key: $EXA_API_KEY`); create keys at https://dashboard.exa.ai/api-keys?utm_source=docs&utm_medium=docs&utm_campaign=markdown_instructions. Every way an agent can get access to Exa is described at https://exa.ai/auth.md.
> Prefer the official SDKs, `exa-py` (`pip install exa-py`) and `exa-js` (`npm install exa-js`); both read `EXA_API_KEY` from the environment.
> Tool-using agents can call Exa without writing code through the hosted MCP server at https://mcp.exa.ai/mcp, or install the Exa agent skill with `npx skills add exa-labs/agent-skills` (skill file: https://exa.ai/docs/skill.md).
> The OpenAPI specs at https://exa.ai/docs/exa-spec.yaml and https://exa.ai/docs/team-management-spec.yaml are the source of truth for request and response schemas.

# Exa Snapshot

> Pin Search and Contents to a stored version of a page at a datetime you choose.

Exa Snapshot keeps stored versions of pages Exa has crawled. Send `snapshotAsOf` to pin the request to a datetime.

Use this to backtest agents, run reproducible evals, and compare earlier versions of docs, pricing pages, policies, and filings.

> [!NOTE]
>
> Exa Snapshot is available on pay as you go at 10 QPS, with a rolling 5-month index window.
> Please reach out to our team for access to our entire historical coverage.
> After 100 requests, [talk to sales](https://exa.ai/contact/sales) to continue. Enterprise plans
> include Snapshot when the contract covers it.
> The Free tier does not include Snapshot, whatever its credit balance. Add a card and buy credits
> on the [Billing dashboard](https://dashboard.exa.ai/billing?utm_source=docs\&utm_medium=docs\&utm_campaign=search_snapshot) to move to a pay-as-you-go plan.

## Search at a datetime

On `/search`, put `snapshotAsOf` inside `contents`.

```python Python theme={null}
from exa_py import Exa

exa = Exa()

result = exa.search(
    "latest stable Python release notes",
    num_results=3,
    contents={
        "snapshot_as_of": "2026-07-01T00:00:00Z",
        "highlights": True,
    },
)

for r in result.results:
    print(r.title, r.url)
```

```javascript JavaScript theme={null}
import Exa from "exa-js";

const exa = new Exa();

const result = await exa.search("latest stable Python release notes", {
  numResults: 3,
  contents: {
    snapshotAsOf: "2026-07-01T00:00:00Z",
    highlights: true
  }
});

for (const r of result.results) {
  console.log(r.title, r.url);
}
```

```bash cURL theme={null}
curl -s -X POST "https://api.exa.ai/search" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $EXA_API_KEY" \
  -d '{
    "query": "latest stable Python release notes",
    "numResults": 3,
    "contents": {
      "snapshotAsOf": "2026-07-01T00:00:00Z",
      "highlights": true
    }
  }'
```

Exa discovers candidate URLs, then keeps only pages with a stored version at or before `snapshotAsOf`.

> [accordion] Example response
>
> ```json theme={null}
> {
>   "requestId": "211fc1f57b87a792de082309ef3bce95",
>   "results": [
>     {
>       "id": "https://docs.python.org/3/whatsnew/changelog.html",
>       "url": "https://docs.python.org/3/whatsnew/changelog.html",
>       "title": "Changelog — Python 3.14.6 documentation",
>       "highlights": [
>         "Changelog — Python 3.14.6 documentation\n...\n## Python 3.14.6 final¶\n...\nRelease date: 2026-06-10"
>       ],
>       "image": "https://docs.python.org/3.14/_images/social_previews/..."
>     },
>     {
>       "id": "https://docs.python.org/3/whatsnew/index.html",
>       "url": "https://docs.python.org/3/whatsnew/index.html",
>       "title": "What's New in Python — Python 3.14.6 documentation",
>       "highlights": ["What's new in Python\n...\n- Python 3.14.6 final\n- Python 3.14.5 final"]
>     },
>     {
>       "id": "https://docs.python.org/3/whatsnew/3.14.html",
>       "url": "https://docs.python.org/3/whatsnew/3.14.html",
>       "title": "What's new in Python 3.14 — Python 3.14.6 documentation",
>       "highlights": ["Python 3.14 is the latest stable release of the Python programming language..."]
>     }
>   ]
> }
> ```

## Pin contents to a datetime

Add `snapshotAsOf` at the top level of a `/contents` request.

```python Python theme={null}
from exa_py import Exa

exa = Exa()

result = exa.get_contents(
    ["https://en.wikipedia.org/wiki/2026"],
    snapshot_as_of="2026-06-01T00:00:00Z",
    text=True,
)

print(result.results[0].text[:300])
```

```javascript JavaScript theme={null}
import Exa from "exa-js";

const exa = new Exa();

const result = await exa.getContents(
  ["https://en.wikipedia.org/wiki/2026"],
  {
    snapshotAsOf: "2026-06-01T00:00:00Z",
    text: true
  }
);

console.log(result.results[0].text.slice(0, 300));
```

```bash cURL theme={null}
curl -s -X POST "https://api.exa.ai/contents" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $EXA_API_KEY" \
  -d '{
    "ids": ["https://en.wikipedia.org/wiki/2026"],
    "snapshotAsOf": "2026-06-01T00:00:00Z",
    "text": true
  }'
```

Exa returns the newest stored version at or before that datetime.

> [accordion] Example response
>
> ```json theme={null}
> {
>   "requestId": "c05151f7df9cd9d8785e0acf0935355d",
>   "results": [
>     {
>       "id": "https://en.wikipedia.org/wiki/2026",
>       "url": "https://en.wikipedia.org/wiki/2026",
>       "title": "2026",
>       "author": null,
>       "text": "2026\n\n2026 (MMXXVI) is the current year, and is a common year starting on Thursday of the Gregorian calendar...",
>       "image": "https://upload.wikimedia.org/wikipedia/commons/thumb/9/93/..."
>     }
>   ],
>   "statuses": [
>     {
>       "id": "https://en.wikipedia.org/wiki/2026",
>       "status": "success",
>       "source": "cached"
>     }
>   ]
> }
> ```

> [!TIP]
>
> IDs with no eligible version are omitted from `results` and reported in `statuses` with
> `"status": "error"` and an `error` object whose `tag` is `CONTENT_NOT_CACHED`:
> `{"status": "error", "error": {"tag": "CONTENT_NOT_CACHED", "httpStatusCode": 404}}`.

## How snapshots work

| Field | Where | Meaning |
| - | - | - |
| `snapshotAsOf` | Request | Cutoff instant as an ISO 8601 date-time (`2026-06-01T00:00:00Z`) or date (`2026-06-01`, midnight UTC). Exa returns the newest stored version at or before this instant. |

For both endpoints:

* Returned page content comes from that stored version.
* Title, author, publication date, text, highlights, and summaries are generated only from that version.
* Pages without an eligible version in the 5-month window are omitted. Please reach out to our team for access to our entire historical coverage.

> [!NOTE]
>
> On Search, the cutoff bounds content, not ranking. Exa still uses current retrieval signals to
> discover candidate URLs. Use the results as evidence bounded by `snapshotAsOf`, not as an exact
> reconstruction of what a search would have ranked at that time.

## Limits and compatibility

> [accordion] Access, rate limit, and lookback
>
> Pay as you go includes 10 QPS and a rolling 5 months of index access. A `snapshotAsOf` older
> than that window is rejected. Please reach out to our team for access to our entire historical coverage.
> After 100 requests, [talk to sales](https://exa.ai/contact/sales)
> to continue.
>
> | Plan | Snapshot access |
> | - | - |
> | Free | Not included. Requests return `403` `SNAPSHOT_NOT_ON_PLAN`, and the Snapshot control on the playground is disabled. Add a card and buy credits on the [Billing dashboard](https://dashboard.exa.ai/billing?utm_source=docs\&utm_medium=docs\&utm_campaign=search_snapshot) to move to a pay-as-you-go plan. |
> | Pay as you go | 100 requests at 10 QPS, up to 10 search results or 1 contents URL per request. Past that, requests return `403` `SNAPSHOT_TRIAL_EXHAUSTED` or `SNAPSHOT_TRIAL_CAP_EXCEEDED`. |
> | Enterprise | Included when your contract covers it; otherwise requests return `403` `SNAPSHOT_NOT_IN_CONTRACT`. [Talk to sales](https://exa.ai/contact/sales) to add it to your contract. |

> [accordion] Historical requests use stored content
>
> Do not combine `snapshotAsOf` with options that can reach the live web or expand to other pages.
> Omit `livecrawl`, `livecrawlTimeout`, `maxAgeHours`, and `subpages` entirely; requests that set
> any of them alongside `snapshotAsOf` are rejected with `INVALID_REQUEST`.

> [accordion] Supported Search requests
>
> Exa Snapshot on Search supports `auto`, `fast`, and `instant`. `deep-lite` is supported on teams
> where it is enabled; [talk to sales](https://exa.ai/contact/sales) to turn it on. It does not
> support `deep` or `deep-reasoning`. It also supports `lite`. `fast` and `instant` are
> [legacy search types](https://exa.ai/docs/search/quickstart.md#legacy-search-types).
>
> Exa Snapshot supports all Search categories through the `category` parameter.

## Common uses

Use Exa Snapshot when the task depends on what Exa had stored by a specific datetime:

* Backtest an agent without exposing it to later page updates.
* Run an evaluation against a repeatable content boundary.
* Compare earlier versions of documentation, pricing, policies, or filings.