VoxCeleb2
Models compare two interview recordings and decide whether the same person speaks in both.
HELM uses VoxCeleb2 test recordings for speaker verification. The release month refers to the original corpus.
- Made by
- Joon Son Chung and colleagues, University of Oxford
- Released
- June 2018
- Kind
- Audio understanding and voice
- Level
- Everyday
- Scored
- Checked automatically
- Answers
- Multiple choice
- Tasks
- All public
- Measures
- Capability
- Language
- Many languages
- Interaction
- Single turn
- Items from
- Real work and use
- Refresh
- New versions replace it
- Run by
- Anyone
- Based on
- VoxCeleb1
- In views
- Audio and speech, Many languages
- Strongest model tested
-
Gemini 2.5 Flash May 2025 #114 on the capability index
- Last model added
- May 2025
- Scores refreshed
- November 20th, 2026
Gemini 2.5 Flash leads with 75.9% among models released by May 2025 Gemini 2.5 Flash leads with 75.9% among models released by May 2025
Best score per entry from AHELM’s leaderboard, read on November 20th, 2026. Where a model ran on its own with several settings, we keep its best. Agents other companies built around a model get a pale shade of that model's lab color. Gray marks blends and entries not in our model list, such as fine-tunes and older models. None of these count toward leaders. Agents built around a model, such as whisper-1 transcription (GPT-4o + whisper-1), are their own entries. We don't count them as the model's score. This view shows only models we track, each at its best.
The newest model with a score here came out in May 2025, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 75.9
Gemini 2.5 Flash
- 75.1
Gemini 2.5 Pro
- 60.1 whisper-1 transcription (GPT-4o + whisper-1)
- 58.5
Gemini 1.5 Pro
- 58.1
Qwen2.5-Omni-7B
- 57.5 gpt-4o-audio-preview-2024-12-17
- 57.0 gpt-4o-audio-preview-2024-10-01
- 54.8 gpt-4o-mini-transcribe transcription (GPT-4o + gpt-4o-mini-transcribe)
- 54.2
Gemini 1.5 Flash
- 53.0
Gemini 2.0 Flash
- 52.9
Gemini 2.0 Flash
- 52.7
Gemini 2.0 Flash-Lite
- 52.4
Gemini 1.5 Pro
- 52.2
Gemini 1.5 Flash
- 52.1 gpt-4o-transcribe transcription (GPT-4o + gpt-4o-transcribe)
- 50.9 gpt-4o-mini-audio-preview-2024-12-17
- 24.0
Qwen2-Audio-7B-Instruct
- 1
Gemini 2.5 Flash May 2025 75.9%
- 2
Gemini 2.5 Pro May 2025 75.1%
- 3 whisper-1 transcription (GPT-4o + whisper-1) Nov 2024 60.1%
- 4
Gemini 1.5 Pro Sep 2024 58.5%
- 5
Qwen2.5-Omni-7B 58.1%
- 6 gpt-4o-audio-preview-2024-12-17 57.5%
- 7 gpt-4o-audio-preview-2024-10-01 57%
- 8 gpt-4o-mini-transcribe transcription (GPT-4o + gpt-4o-mini-transcribe) Nov 2024 54.8%
- 9
Gemini 1.5 Flash Sep 2024 54.2%
- 10
Gemini 2.0 Flash Dec 2024 53%
- 11
Gemini 2.0 Flash Feb 2025 52.9%
- 12
Gemini 2.0 Flash-Lite 52.7%
- 13
Gemini 1.5 Pro May 2024 52.4%
- 14
Gemini 1.5 Flash May 2024 52.2%
- 15 gpt-4o-transcribe transcription (GPT-4o + gpt-4o-transcribe) Nov 2024 52.1%
- 16 gpt-4o-mini-audio-preview-2024-12-17 50.9%
- 17
Qwen2-Audio-7B-Instruct 24%
Source: AHELM’s leaderboard
- 1
Gemini 2.5 Flash May 2025 75.9%
- 2
Gemini 2.5 Pro May 2025 75.1%
- 3
Gemini 1.5 Pro Sep 2024 58.5%
- 4
Qwen2.5-Omni-7B 58.1%
- 5
Gemini 1.5 Flash Sep 2024 54.2%
- 6
Gemini 2.0 Flash Dec 2024 53%
- 7
Gemini 2.0 Flash Feb 2025 52.9%
- 8
Gemini 2.0 Flash-Lite 52.7%
- 9
Gemini 1.5 Pro May 2024 52.4%
- 10
Gemini 1.5 Flash May 2024 52.2%
- 11
Qwen2-Audio-7B-Instruct 24%
Source: AHELM’s leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 75.9% | May 2025 | Official leaderboard | |
| 2 | | 75.1% | May 2025 | Official leaderboard | |
| 3 | whisper-1 transcription (GPT-4o + whisper-1) Nov 2024 | 60.1% | Nov 2024 | Official leaderboard | |
| 4 | | 58.5% | Sep 2024 | Official leaderboard | |
| 5 | | 58.1% | Mar 2025 | Official leaderboard | |
| 6 | gpt-4o-audio-preview-2024-12-17 | 57.5% | Official leaderboard | ||
| 7 | gpt-4o-audio-preview-2024-10-01 | 57% | Official leaderboard | ||
| 8 | gpt-4o-mini-transcribe transcription (GPT-4o + gpt-4o-mini-transcribe) Nov 2024 | 54.8% | Nov 2024 | Official leaderboard | |
| 9 | | 54.2% | Aug 2024 | Official leaderboard | |
| 10 | | 53% | Dec 2024 | Official leaderboard | |
| 11 | | 52.9% | Feb 2025 | Official leaderboard | |
| 12 | | 52.7% | Feb 2025 | Official leaderboard | |
| 13 | | 52.4% | May 2024 | Official leaderboard | |
| 14 | | 52.2% | May 2024 | Official leaderboard | |
| 15 | gpt-4o-transcribe transcription (GPT-4o + gpt-4o-transcribe) Nov 2024 | 52.1% | Nov 2024 | Official leaderboard | |
| 16 | gpt-4o-mini-audio-preview-2024-12-17 | 50.9% | Official leaderboard | ||
| 17 | | 24% | Aug 2024 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 75.9% | May 2025 | Official leaderboard | |
| 2 | | 75.1% | May 2025 | Official leaderboard | |
| 3 | | 58.5% | Sep 2024 | Official leaderboard | |
| 4 | | 58.1% | Mar 2025 | Official leaderboard | |
| 5 | | 54.2% | Aug 2024 | Official leaderboard | |
| 6 | | 53% | Dec 2024 | Official leaderboard | |
| 7 | | 52.9% | Feb 2025 | Official leaderboard | |
| 8 | | 52.7% | Feb 2025 | Official leaderboard | |
| 9 | | 52.4% | May 2024 | Official leaderboard | |
| 10 | | 52.2% | May 2024 | Official leaderboard | |
| 11 | | 24% | Aug 2024 | Official leaderboard |
5 models have held number one on AHELM’s leaderboard since 2023
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: AHELM’s leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: AHELM’s leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Audio and video
- Video-MMMU In use 300 expert-level lecture videos with questions testing perception, comprehension and adaptation of knowledge. Accuracy.
Qwen 3.8 Max 88.7%
- LVBench In use Hours-long video understanding with multiple-choice questions on reasoning, grounding and summarization.
Gemini 4 Argon 91.7%
- MMVU In use 3,000 expert-annotated questions on videos across 27 disciplines. Multiple-choice and open-ended accuracy.
Qwen 3.8 Max 82.4%
- MLVU In use Multi-task long video understanding (minutes to hours) with multiple-choice and generation tasks. Accuracy.
Qwen 3.8 Max 90.8%
- FLEURS In use Models transcribe read speech in 102 languages built from FLoRes-101 sentences. Model cards usually report word error rate, so lower scores are better.
Phi-4-multimodal 4.00%
- VoiceBench In use Spoken instructions with speaker, environment and content variation. Voice assistant responses scored by GPT judge or rules.
Qwen3.5-Omni-Plus 93.1%
Data audit
9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 18 months ago, in May 2025, for Gemini 2.5 Flash. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
- Pass Independent runs Scores from the official leaderboard (17 entries).
- Pass Models scored 11 models we track, plus 6 other entries on its leaderboard.
- Warning Newest score The newest came 18 months ago, in May 2025, for Gemini 2.5 Flash.
- Note Frontier models None of the top 10 has a score. The strongest tested is Gemini 2.5 Flash (#114).
- Pass Headroom The best is 75.9%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on AHELM: 14 rows for models we track and 3 other entries at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.