GRADIUMTEXT-TO-SPEECH11,934 SAMPLES / 30 DAYSLAST RUN NOV 21, 2026, 03:30 UTC
Default text-to-speech benchmarks
Default, hosted by Gradium, measures mean 62 ms time to first audio (2nd of 32) and 1.7% word error rate (16th of 32) among TTS systems. Results cover the last 30 days. Coval benchmarks hosted endpoints daily using a consistent evaluation methodology. Last measured .
Coval has active text-to-speech measurements for Default, created by Gradium and served through Gradium's own API.
- Time to First Audio#2 / 32
- 62ms
- TTFA Network Roundtrip#3 / 32
- 60ms
- TTFA Leading Silence#1 / 32
- 3ms
- Word Error Rate#16 / 32
- 1.7%
Overview
Default is tested every day on a fixed set of text prompts, measuring how quickly audible speech starts and how intelligible the result is.
How Default ranks
Full TTS dashboard- #7Qwen3 TTS 1.7bDedicated inference. Shared endpoints serve many customers on the same infrastructure, while dedicated endpoints run on hardware reserved for a single customer.105 ms
Show all 32 modelsShow fewer
Show all 32 modelsShow fewer
- #17Qwen3 TTS 1.7bDedicated inference. Shared endpoints serve many customers on the same infrastructure, while dedicated endpoints run on hardware reserved for a single customer.1.7%
Latency vs accuracy
Where the errors come from
- Default1.7%
Averages and tail latency
Averages hide slow outliers — these are the distributions behind each figure.
- Time to First Audiop50 53 ms · Max 278 ms
- TTFA Network Roundtripp50 52 ms · Max 239 ms
- TTFA Leading Silencep50 0 ms · Max 70 ms
- Defaultp50 0.0% · Max 55.6%
| Metric | Average | p25 | p50 | p75 | p90 | p95 | Max | Samples |
|---|---|---|---|---|---|---|---|---|
| Time to First Audio | 62 ms | 49 ms | 53 ms | 70 ms | 93 ms | 109 ms | 278 ms | 2,876 |
| TTFA Network Roundtrip | 60 ms | 48 ms | 52 ms | 65 ms | 84 ms | 100 ms | 239 ms | 2,876 |
| TTFA Leading Silence | 3 ms | 0 ms | 0 ms | 0 ms | 3 ms | 33 ms | 70 ms | 2,876 |
| Word Error Rate | 1.7% | 0.0% | 0.0% | 2.8% | 9.3% | 13.2% | 55.6% | 2,824 |
Last 30 days
Daily medians from the same measurement runs · gaps are days without qualifying runs.
How fast is Default?
On Gradium, Default measures mean 62 ms time to first audio (2nd of 32). Last measured 2026-11-21.
How accurate is Default?
On Gradium, Default measures 1.7% word error rate (16th of 32). Last measured 2026-11-21.
Who hosts Default?
Default is created by Gradium and served by Gradium. Coval measures each hosted endpoint separately.
Results are re-measured daily using fixed datasets and reported over a rolling 30-day window. Same datasets, prompts and metric definitions for every model, measured by Coval’s open-source runner. Full methodology on the overview.
Evaluate your own voice agent
Use Coval to test your production configuration, prompts and calls—not only the public benchmark endpoints.