GAIK LA-murre dialect test
Models transcribe 583 Finnish dialect interview segments from ten localities. The reference preserves dialect forms, so standard Finnish spellings can count as errors.
GAIK reports 18 models across 20 scored configurations. Separate model judges score changes in meaning.
- Released
- October 2026
- Kind
- Speech recognition and translation
- Level
- Everyday
- Scored
- Checked automatically
- Answers
- Open-ended
- Tasks
- All public
- Measures
- Capability
- Language
- Finnish
- Interaction
- Single turn
- Items from
- Real work and use
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Audio and speech
- Strongest model tested
-
Gemini 3.5 Flash #44 on the capability index
- Last model added
- May 2026
- Scores refreshed
- November 20th, 2026
Finnish-NLP whisper-large-finnish-v3 tops the leaderboard with 51.6% Gemini 3.5 Flash leads with 60.2% among models released by May 2026
Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. 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 Gemini 3.5 Flash, at 60.2%. This view shows only models we track, each at its best.
The newest model with a score here came out in May 2026, 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
- 1 Finnish-NLP whisper-large-finnish-v3 51.6%
- 2 Ours: main model (parl) 57.4%
- 3 Ours: la-murre version 57.4%
- 4 Ours: all-skn version 57.8%
- 5 Ours: round 1 58%
- 6 Ours: first model (v1.0) 58.1%
- 7 Ours: timestamps version 58.1%
- 8 RASMUS Finnish-ASR-Canary-v2, greedy 58.7%
- 9
Gemini 3.5 Flash API, asked to write as spoken 60.2%
- 10 Whisper large-v3 62.5%
- 11 Whisper large-v3-turbo 62.7%
- 12 RASMUS Finnish-ASR-Canary-v2 + KenLM 62.7%
- 13 Orukeet (Parakeet v3 fine-tune) 64%
- 14 NVIDIA Canary 1B v2 65.6%
- 15 NVIDIA Nemotron 3.5 Transcribe 66.4%
- 16 NVIDIA Parakeet TDT 0.6B v3 66.7%
- 17 Finnish-NLP wav2vec2 XLS-R 1B (no LM) 67.3%
- 18 OpenAI gpt-4o-transcribe (API, via Azure) API, via Azure 72.6%
- 19
Qwen3-ASR-1.7B 72.9%
Source: Official GAIK LA-murre dialect test leaderboard
- 1
Gemini 3.5 Flash API, asked to write as spoken 60.2%
- 2 Whisper large-v3 62.5%
- 3 Whisper large-v3-turbo 62.7%
- 4
Qwen3-ASR-1.7B 72.9%
Source: Official GAIK LA-murre dialect test leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Finnish-NLP whisper-large-finnish-v3 | 51.6% | Official leaderboard | ||
| 2 | Ours: main model (parl) | 57.4% | Official leaderboard | ||
| 3 | Ours: la-murre version | 57.4% | Official leaderboard | ||
| 4 | Ours: all-skn version | 57.8% | Official leaderboard | ||
| 5 | Ours: round 1 | 58% | Official leaderboard | ||
| 6 | Ours: first model (v1.0) | 58.1% | Official leaderboard | ||
| 7 | Ours: timestamps version | 58.1% | Official leaderboard | ||
| 8 | RASMUS Finnish-ASR-Canary-v2, greedy | 58.7% | Official leaderboard | ||
| 9 | | 60.2% | API, asked to write as spoken | May 2026 | Official leaderboard |
| 10 | Whisper large-v3 | 62.5% | Nov 2023 | Official leaderboard | |
| 11 | Whisper large-v3-turbo | 62.7% | Oct 2024 | Official leaderboard | |
| 12 | RASMUS Finnish-ASR-Canary-v2 + KenLM | 62.7% | Official leaderboard | ||
| 13 | Orukeet (Parakeet v3 fine-tune) | 64% | Official leaderboard | ||
| 14 | NVIDIA Canary 1B v2 | 65.6% | Official leaderboard | ||
| 15 | NVIDIA Nemotron 3.5 Transcribe | 66.4% | Official leaderboard | ||
| 16 | NVIDIA Parakeet TDT 0.6B v3 | 66.7% | Official leaderboard | ||
| 17 | Finnish-NLP wav2vec2 XLS-R 1B (no LM) | 67.3% | Official leaderboard | ||
| 18 | OpenAI gpt-4o-transcribe (API, via Azure) | 72.6% | API, via Azure | Official leaderboard | |
| 19 | | 72.9% | Jan 2026 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 60.2% | API, asked to write as spoken | May 2026 | Official leaderboard |
| 2 | Whisper large-v3 | 62.5% | Nov 2023 | Official leaderboard | |
| 3 | Whisper large-v3-turbo | 62.7% | Oct 2024 | Official leaderboard | |
| 4 | | 72.9% | Jan 2026 | Official leaderboard |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official GAIK LA-murre dialect test 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 GAIK LA-murre dialect test 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 6 months ago, in May 2026, for Gemini 3.5 Flash. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
- Pass Independent runs Scores from the official leaderboard (19 entries).
- Pass Models scored 4 models we track, plus 15 other entries on its leaderboard.
- Warning Newest score The newest came 6 months ago, in May 2026, for Gemini 3.5 Flash.
- Note Frontier models None of the top 10 has a score. The strongest tested is Gemini 3.5 Flash (#44).
- Note Headroom Lower is better, so there’s no ceiling to reach.
Leaderboard import 3 pass
- Pass Importer We read the board on github.com: 5 rows for models we track and 15 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.