HALAS
Models transcribe earnings call segments with human annotations of hallucinations. NVIDIA measures word error and segments with inserted text.
- Made by
- Mateusz Barański and colleagues
- Released
- June 2026
- Kind
- Speech recognition and translation
- Level
- Everyday
- Scored
- Checked automatically
- Answers
- Open-ended
- Tasks
- All public
- Measures
- Propensity
- Behavior
- Hallucination
- Language
- English
- Interaction
- Single turn
- Items from
- Real work and use
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Audio and speech
- Last model added
- October 2024
- Scores refreshed
- November 20th, 2026
Crisper Whisper tops the leaderboard with 21.4% Whisper large-v3 leads with 23.8%
Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is Whisper large-v3, at 23.8%. This view shows only models we track, each at its best.
- 1 Crisper Whisper 21.4%
- 2 Whisper large-v3 23.8%
- 3 Whisper large-v3-turbo 29.4%
- 4 Canary-1B-Flash 30.4%
- 5 Canary-1B 33.6%
- 6 Parakeet-TDT version 2 33.7%
- 7 Whisper large v2 43.8%
Source: The HALAS paper
- 1 Whisper large-v3 23.8%
- 2 Whisper large-v3-turbo 29.4%
Source: The HALAS paper
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Crisper Whisper | 21.4% | Official leaderboard | ||
| 2 | Whisper large-v3 | 23.8% | Nov 2023 | Official leaderboard | |
| 3 | Whisper large-v3-turbo | 29.4% | Oct 2024 | Official leaderboard | |
| 4 | Canary-1B-Flash | 30.4% | Official leaderboard | ||
| 5 | Canary-1B | 33.6% | Official leaderboard | ||
| 6 | Parakeet-TDT version 2 | 33.7% | Official leaderboard | ||
| 7 | Whisper large v2 | 43.8% | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Whisper large-v3 | 23.8% | Nov 2023 | Official leaderboard | |
| 2 | Whisper large-v3-turbo | 29.4% | Oct 2024 | Official leaderboard |
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official HALAS 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
8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 2 years ago, in October 2024, for Whisper large-v3-turbo. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
- Pass Independent runs Scores from the benchmark’s paper (7 entries).
- Pass Models scored 2 models we track, plus 5 other entries on its leaderboard.
- Warning Newest score The newest came 2 years ago, in October 2024, for Whisper large-v3-turbo.
- Note Frontier models None of its models are on the capabilities index.
- Note Headroom Lower is better, so there’s no ceiling to reach.
Leaderboard import 3 pass
- Pass Importer We read the board on arxiv.org: 2 rows for models we track and 5 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.