← Speech-to-Text

Nari

Qwen3-ASR Fast

Measured 2026-09-13

Run Qwen3-ASR Fast on Speko →

One API, every model on this board.

WER · batch

3.3%

2.0–4.8%

WER · streaming

Read speech

3.83%

Conversation

11.99%

Details

Read speech

n=1541 · FLEURS test+dev+train · vendor-direct

Conversation

9.9–14.3% · EdAcc n=283 · 60 speakers, 2 passes, 2026-10-01

Finalize

33ms

p50–p90 · US East · n=60 · vendor-direct · pre-close (30/30 flush-forced) · re-measured 2026-09-22 · ±9%

Commit-to-final: turn detection is off, so the probe commits at the measured speech end. Singapore reads 196ms.

Time to first token

1.74s

p50–p90 · US East · n=60 · vendor-direct · re-measured 2026-09-22 · ±0%

Cost / min

$0.002

Vendor list $0.12 per input audio hour, Partner tier (docs.narilabs.com/models-and-pricing, read 2026-09-13). A free tier of the same model is listed at $0.

Share this result

Embed the live badge

Speko stt rank
Markdown
[![Speko stt rank](https://benchmarks.speko.ai/badge/stt/nari-qwen3-asr-fast.svg)](https://benchmarks.speko.ai/stt/nari-qwen3-asr-fast)
HTML
<a href="https://benchmarks.speko.ai/stt/nari-qwen3-asr-fast"><img src="https://benchmarks.speko.ai/badge/stt/nari-qwen3-asr-fast.svg" alt="Speko stt rank"></a>
URL
https://benchmarks.speko.ai/stt/nari-qwen3-asr-fast