OpenAI GPT-Realtime-2
Listed Proprietary · OpenAI
5.1
/100
OpenAI GPT-Realtime-2 is a proprietary text-to-speech (TTS) API from OpenAI, measured by one independent benchmark, Artificial Analysis.
Artificial Analysis
Fish Audio OpenAudio S1 Mini Fish Audio · text-to-speech (TTS) API
ListedSpeech Arena Elo 1,042 on Artificial Analysis, latency not measured; ranked 76th (tied) of 112 in the TTS for Voice Agents 2026 index.
Benchmark summary: 5.1/100, ranked 76th (tied) of 112 in the TTS for Voice Agents 2026 index (Listed). Strongest criterion: price, 1.8 of 5 points.
Fish Audio OpenAudio S1 Mini is an open-weights text-to-speech (TTS) API from Fish Audio, measured by one independent benchmark, Artificial Analysis.
Open weights Spanish German French Portuguese Japanese Chinese Arabic Artificial Analysis Fish Audio Listed
Overall score
5.1 /100
Score breakdown
- Real-time speed 0 /60
- Voice quality 3.3 /20
- Accuracy 0 /15
- Price 1.8 /5
Data checked on September 30, 2026
Updated after each benchmark capture
Every point comes from a public benchmark: sources · methodology
Fish Audio OpenAudio S1 Mini ranks 76th of 112 in the 2026 text-to-speech (TTS) API benchmark, with a score of 5.1/100.
Real-time latency
Neither Coval nor Speko measures Fish Audio OpenAudio S1 Mini's latency, so it scores 0 of 60 on real-time speed and does not appear in the real-time ranking.
Accuracy (word error rate)
No accuracy measurement: Coval does not measure its word error rate and Speko does not rate its pronunciation robustness, so accuracy counts 0 of 15.
Voice quality
In the Artificial Analysis Speech Arena (capture of October 1, 2026), Fish Audio OpenAudio S1 Mini has an Elo of 1,042 (±20, 95% confidence) from 1,682 blind comparisons, 62nd highest of the 92 models Artificial Analysis rates.
Price
Artificial Analysis lists Fish Audio OpenAudio S1 Mini at $15 per 1M characters (19th lowest of the 78 models Artificial Analysis lists a price for).
Languages tested
Tested in Spanish, German, French, Portuguese, Japanese, Chinese and Arabic by Artificial Analysis or Speko. A language missing here was not tested, which does not mean it is unsupported.
Benchmark measurements
Artificial Analysis Speech Arena
- Speech Arena Elo (blind listener preference)
- 1,042
- Elo 95% interval (±)
- 20
- Arena rank
- 62
- Arena rank range
- 52-71
- Arena appearances
- 1,682
- Voices tested
- 8
- Price per 1M characters
- 15 USD
- Released
- Jun 2025
- Artificial Analysis model
- OpenAudio S1 Mini
- Artificial Analysis leaderboard
- artificialanalysis.ai
Not collected
- Speed (Coval time to first audio) — not measured by Coval (capture of 2026-10-01, 30-day window); counts 0
- Speed (Speko time to first audio) — not measured by Speko (capture of 2026-10-01); counts 0
- Quality (Speko naturalness) — not measured by Speko (capture of 2026-10-01); counts 0
- Accuracy (Coval word error rate) — not measured by Coval (capture of 2026-10-01, 30-day window); counts 0
- Accuracy (Speko robustness) — not measured by Speko (capture of 2026-10-01); counts 0
- Price (Speko cost per 1M characters) — not measured by Speko (capture of 2026-10-01); counts 0
Similar TTS models
ElevenLabs Flash v2
Listed ElevenLabs
5.2
/100
ElevenLabs Flash v2 is a text-to-speech (TTS) API from ElevenLabs, measured by one independent benchmark, Speko.
Speko
MiniMax T2A-01-HD
Listed Proprietary · MiniMax
5.2
/100
MiniMax T2A-01-HD is a proprietary text-to-speech (TTS) API from MiniMax, measured by one independent benchmark, Artificial Analysis.
Artificial Analysis