---
license: cc-by-nc-4.0
extra_gated_heading: "Acknowledge license to access the dataset"
extra_gated_description: "To request access, please review and accept the terms below. Our team may take 2–3 business days to process your request."
extra_gated_button_content: "Request access"
extra_gated_prompt: "This data is licensed under CC-BY-NC. Those terms notwithstanding, by downloading this data you agree that:
You will use this data only for the purposes of testing transcription quality; 
You will not share or redistribute this data; and
At the earliest of your completion of use of the data for the purpose described above, or 10 days, you will securely delete the downloaded data and any copies you may have created or allowed to be created."
extra_gated_fields:
  Country: country
  Company/Affiliation: text
  Role: text
  I agree to use this dataset for non-commercial purposes only: checkbox
---
# Dataset Card for μ-Bench ([Leaderboard](https://research.sierra.ai/mubench) | [Code](https://github.com/sierra-research/mu-bench))

<!-- Provide a quick summary of the dataset. -->

μ-Bench is a multilingual transcription benchmark built from real customer-service phone conversations. 

## Dataset Details

### Dataset Description

<!-- Provide a longer summary of what this dataset is. -->
Most public ASR benchmarks are either English-only or built from read speech in quiet studios. μ-Bench fills that gap: real phone-call audio, five languages, and metrics that go beyond Word Error Rate to distinguish meaning-changing errors from surface-level ones.

The calls are scripted interactions with an AI banking agent built on Sierra's voice platform — checking card status, confirming case codes, providing personal details, disputing transactions, and requesting credit-limit increases.

Callers used their own phones from their own environments, producing realistic background noise, disfluencies, emotional variation, interruptions, and diverse speaking styles.

The call audio is processed and clipped into individual user utterances. 

## Dataset Structure

<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->

The dataset contains 250 conversations totaling 4,270 utterances (~5.1 hours of audio):

| Locale | Language           | Utterances |
|--------|--------------------|------------|
| en-US  | English (US)       | 817        |
| es-MX  | Spanish (Mexico)   | 792        |
| tr-TR  | Turkish            | 846        |
| vi-VN  | Vietnamese         | 975        |
| zh-CN  | Chinese (Mandarin) | 840        |

Audio files follow:

`<locale>/conv-<N>-turn-<M>.wav`

Each utterance has a row in `metadata.jsonl`:

| Field             | Type   | Description |
|-------------------|--------|------------|
| file_name         | string | Relative audio path |
| locale            | string | BCP-47 locale tag |
| conversation_id   | int    | Conversation identifier |
| turn_index        | int    | Turn index (0-based) |
| duration_sec      | float  | Duration in seconds |
| transcript        | string | Clean verbatim transcript |

There is a single split. Evaluation uses the full dataset with conversation-level bootstrap resampling.

## Evaluation Artifacts

In addition to the dataset, the prompts used for LLM-based normalization and error classification (see our Github for more details) are also available for download as part of this repository.

## More Information

- Leaderboard: https://research.sierra.ai/mubench
- Code & submissions: https://github.com/sierra-research/mu-bench

## Dataset Card Contact

soham@sierra.ai