---
title: DigitalOcean Evaluations Limits
description: Limits and known issues for Evaluations.
product: Evaluations
url: https://docs.digitalocean.com/products/evaluations/details/limits/
last_updated: "2026-10-12"
---

> **For AI agents:** The documentation index is at [https://docs.digitalocean.com/llms.txt](https://docs.digitalocean.com/llms.txt). Markdown versions of pages use the same URL with `index.html.md` in place of the HTML page (for example, append `index.html.md` to the directory path instead of opening the HTML document).

# DigitalOcean Evaluations Limits

DigitalOcean Evaluations measure model, Inference Router, and agent performance using dataset evaluations and simulations.

## Limits

- Each evaluation can use up to 10 selected metrics, including custom metrics.
- Access to commercial models from OpenAI and Anthropic as candidate or judge models requires a tier 2 account. To move to tier 2 and unlock model access, complete a [prepayment](https://cloud.digitalocean.com/limits?i=b59231) in the DigitalOcean Control Panel.

### Dataset Evaluation Limits

- Each dataset can contain up to 1,000 rows. For multi-turn datasets, the limit is 1,000 conversations.
- Each dataset can be up to 1 GB, regardless of team tier.
- You cannot rename datasets after creation.
- In the Control Panel, custom metrics are available only for single-turn dataset evaluations.

### Simulation Limits (public)

- Each scenario set can contain up to 10 scenarios. Generating a scenario set from a goal creates from 1 to 10 scenarios.
- Each uploaded scenario set file can be up to 5 MB.
- Scenario set names can be up to 255 characters.
- The `max_turns` value for an uploaded scenario must be from `1` through `10`.
- Each scenario can have up to 10 tags. Each tag can be up to 64 characters and can contain only letters, numbers, hyphens (`-`), and underscores (`_`).
- Each simulation run can use up to 10 turns per scenario and up to five episodes per scenario.
- You cannot change the scenarios in, or delete, a scenario set while any simulation run references it. To change the scenarios, duplicate the set and edit the copy.
- You can only delete a simulation run after it reaches a terminal status: succeeded, partially successful, failed, or cancelled.
- In the Control Panel, simulations use only the built-in multi-turn evaluation metrics and do not offer custom metrics. When you create a simulation run using the API, you can also use custom metrics.

### Preset and Custom Metric Limits

- Each team can create up to 100 evaluation presets.
- Each team can create up to 100 custom metrics.
- Each custom metric must contain fewer than 8,192 characters.
- You can only delete custom metrics from the evaluation creation page.

To request a higher preset or custom metric limit, contact [support](https://cloudsupport.digitalocean.com).