File size: 11,762 Bytes
d02cb2a 9a88e93 d02cb2a 9a88e93 d02cb2a 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 50ae038 9a88e93 d02cb2a | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 | <!doctype html>
<html lang="en">
<head>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width, initial-scale=1" />
<meta
name="description"
content="Ringg Parrot STT V1 is a proprietary Hindi-English code-mixed ASR system for real-time business voice applications."
/>
<title>Ringg Parrot STT V1</title>
<link rel="stylesheet" href="style.css" />
</head>
<body>
<header class="site-header">
<nav class="nav shell" aria-label="Primary navigation">
<a class="brand" href="#top" aria-label="Ringg Parrot STT V1 home">
<span class="brand-mark">R</span>
<span>Ringg Parrot STT</span>
</a>
<div class="nav-links">
<a href="#benchmarks">Benchmarks</a>
<a href="#integration">Integration</a>
<a href="#access">Access</a>
</div>
<a class="nav-cta" href="https://ringg.ai/dashboard/stt" target="_blank" rel="noreferrer">
Open Playground
</a>
</nav>
<section class="hero shell" id="top">
<div class="hero-copy">
<div class="tag-row">
<span class="pill">Proprietary ASR</span>
<span>Hindi-English code-mixed speech</span>
</div>
<h1>Production-ready speech-to-text for Hindi-English voice workflows.</h1>
<p class="lead">
Ringg Parrot STT V1 is built for real-time voice products, AI agents,
contact centers, and business transcription workflows that need
reliable Hindi, English, and code-mixed recognition.
</p>
<div class="cta-row">
<a class="button primary" href="https://ringg.ai/dashboard/stt" target="_blank" rel="noreferrer">
Try the Playground
</a>
<a class="button secondary" href="https://pypi.org/project/ringglabs/" target="_blank" rel="noreferrer">
View Python SDK
</a>
</div>
<div class="proof-row" aria-label="Product highlights">
<span>60-80ms streaming latency</span>
<span>Pipecat VAD events</span>
<span>Private model access</span>
</div>
</div>
<aside class="hero-panel" aria-label="Product summary">
<div class="panel-header">
<span>Streaming STT</span>
<strong>V1</strong>
</div>
<div class="signal-card">
<span class="signal-label">Typical latency</span>
<strong>60-80ms</strong>
<p>Designed for low-latency streaming voice experiences.</p>
</div>
<div class="panel-list">
<div>
<span>Languages</span>
<strong>Hindi, English, code-mix</strong>
</div>
<div>
<span>Integration</span>
<strong>SDK and Pipecat-ready events</strong>
</div>
<div>
<span>Access</span>
<strong>Hosted evaluation and commercial access</strong>
</div>
</div>
</aside>
</section>
</header>
<main class="shell">
<section class="metric-strip" aria-label="Key product metrics">
<article class="metric-card">
<span class="metric-value">60-80ms</span>
<span class="metric-label">Typical streaming latency</span>
</article>
<article class="metric-card">
<span class="metric-value">Hindi + English</span>
<span class="metric-label">Code-mixed speech support</span>
</article>
<article class="metric-card">
<span class="metric-value">Proprietary</span>
<span class="metric-label">Private model and implementation</span>
</article>
</section>
<section class="section split" id="access">
<div class="section-copy">
<p class="section-kicker">Access</p>
<h2>Evaluate in the playground. Contact RinggAI for production access.</h2>
<p>
This Space provides product information for Ringg Parrot STT V1. The
model weights, training code, and internal implementation are not open
sourced.
</p>
</div>
<div class="info-card">
<ul class="clean-list">
<li>Playground access is available at ringg.ai.</li>
<li>Model weights are not available for download from this Space.</li>
<li>Production and commercial access requires RinggAI approval.</li>
</ul>
<a class="text-link" href="mailto:sales@ringg.ai">Contact sales@ringg.ai</a>
</div>
</section>
<section class="section split" id="integration">
<div class="section-copy">
<p class="section-kicker">SDK and Integration</p>
<h2>Integrate with voice-agent and real-time audio pipelines.</h2>
<p>
The Ringg SDK helps developers connect Ringg STT into application
workflows. Ringg Parrot STT V1 is highly compatible with Pipecat
toolkit using built-in VAD events.
</p>
</div>
<div class="info-card">
<ul class="clean-list">
<li>Python SDK is available through the ringglabs package on PyPI.</li>
<li>Built for low-latency streaming speech recognition.</li>
<li>Supports modern voice-agent orchestration patterns.</li>
</ul>
<a class="text-link" href="https://pypi.org/project/ringglabs/" target="_blank" rel="noreferrer">
View ringglabs on PyPI
</a>
</div>
</section>
<section class="section" id="benchmarks">
<div class="section-heading">
<p class="section-kicker">Benchmarks</p>
<h2>WER comparison across ASR benchmark datasets.</h2>
<p>
WER stands for Word Error Rate. Lower values indicate better
transcription accuracy. The lowest WER in each row is highlighted.
</p>
</div>
<div class="benchmark-grid">
<div class="table-card">
<div class="table-title">
<h3>Original WER</h3>
<span>Lower is better</span>
</div>
<div class="table-wrap">
<table>
<thead>
<tr>
<th>Dataset</th>
<th>Ringg</th>
<th>ElevenLabs</th>
<th>Deepgram</th>
<th>Sarvam</th>
</tr>
</thead>
<tbody>
<tr>
<td>indictts</td>
<td><strong>11.58</strong></td>
<td>16.06</td>
<td>13.65</td>
<td>15.37</td>
</tr>
<tr>
<td>commonvoice</td>
<td><strong>14.30</strong></td>
<td>16.59</td>
<td>20.04</td>
<td>18.21</td>
</tr>
<tr>
<td>fleurs</td>
<td>15.20</td>
<td><strong>11.99</strong></td>
<td>17.14</td>
<td>16.00</td>
</tr>
<tr>
<td>kathbath</td>
<td><strong>11.78</strong></td>
<td>13.24</td>
<td>15.93</td>
<td>17.53</td>
</tr>
<tr>
<td>kathbath_noisy</td>
<td><strong>13.09</strong></td>
<td>13.14</td>
<td>17.44</td>
<td>16.19</td>
</tr>
<tr>
<td>mucs</td>
<td>14.55</td>
<td><strong>11.69</strong></td>
<td>21.97</td>
<td>16.72</td>
</tr>
<tr>
<td>Overall WER</td>
<td>13.79</td>
<td><strong>13.00</strong></td>
<td>19.23</td>
<td>16.72</td>
</tr>
</tbody>
</table>
</div>
</div>
<div class="table-card">
<div class="table-title">
<h3>Normalized WER</h3>
<span>Lower is better</span>
</div>
<div class="table-wrap">
<table>
<thead>
<tr>
<th>Dataset</th>
<th>Ringg</th>
<th>ElevenLabs</th>
<th>Deepgram</th>
<th>Sarvam</th>
</tr>
</thead>
<tbody>
<tr>
<td>indictts</td>
<td><strong>3.94</strong></td>
<td>8.52</td>
<td>6.93</td>
<td>7.84</td>
</tr>
<tr>
<td>commonvoice</td>
<td><strong>6.37</strong></td>
<td>13.02</td>
<td>14.88</td>
<td>13.06</td>
</tr>
<tr>
<td>fleurs</td>
<td>9.73</td>
<td><strong>7.67</strong></td>
<td>11.35</td>
<td>9.54</td>
</tr>
<tr>
<td>kathbath</td>
<td><strong>7.15</strong></td>
<td>10.15</td>
<td>11.38</td>
<td>10.41</td>
</tr>
<tr>
<td>kathbath_noisy</td>
<td><strong>8.37</strong></td>
<td>10.01</td>
<td>12.98</td>
<td>11.78</td>
</tr>
<tr>
<td>mucs</td>
<td><strong>6.28</strong></td>
<td>6.75</td>
<td>12.07</td>
<td>7.58</td>
</tr>
<tr>
<td>Overall WER</td>
<td><strong>7.27</strong></td>
<td>8.94</td>
<td>12.36</td>
<td>9.76</td>
</tr>
</tbody>
</table>
</div>
</div>
</div>
</section>
<section class="section feature-grid">
<article class="feature-card">
<h2>Features</h2>
<ul class="clean-list">
<li>Hindi-English code-mixed speech recognition.</li>
<li>Real-time streaming transcription.</li>
<li>File-based transcription for common audio formats.</li>
<li>Low-latency inference for voice products.</li>
</ul>
</article>
<article class="feature-card">
<h2>Supported Inputs</h2>
<ul class="clean-list">
<li>Hindi, English, and code-mixed speech.</li>
<li>Clear audio with minimal background noise.</li>
<li>16kHz or higher sample rate recommended.</li>
<li>WAV, MP3, FLAC, M4A, OGG, and OPUS.</li>
</ul>
</article>
<article class="feature-card">
<h2>Use Cases</h2>
<ul class="clean-list">
<li>Voice assistants and AI agents.</li>
<li>Contact center transcription.</li>
<li>Meeting and conversation intelligence.</li>
<li>Voice search, subtitling, and accessibility workflows.</li>
</ul>
</article>
<article class="feature-card">
<h2>Limitations</h2>
<ul class="clean-list">
<li>Accuracy may vary with noisy or low-quality audio.</li>
<li>Overlapping speakers and dialect variation can affect quality.</li>
<li>Very long files or unsupported encodings may require preprocessing.</li>
<li>The hosted demo may differ from production deployment settings.</li>
</ul>
</article>
</section>
<section class="section split">
<div class="section-copy">
<p class="section-kicker">Benchmark Dataset</p>
<h2>Released benchmark data and ASR transcriptions.</h2>
</div>
<div class="info-card">
<p>
RinggAI has released the ASR Benchmarking Open-Source Dataset, which
includes benchmark audio/data and transcriptions generated by Ringg,
ElevenLabs, Deepgram, and Sarvam.
</p>
</div>
</section>
<section class="section split">
<div class="section-copy">
<p class="section-kicker">Privacy and Data Notice</p>
<h2>Review deployment terms before using sensitive data.</h2>
</div>
<div class="info-card">
<p>
Audio handling may depend on the selected deployment, integration, and
commercial terms. Review RinggAI privacy terms and deployment
documentation before using the service with sensitive, regulated, or
personally identifiable data.
</p>
</div>
</section>
</main>
<footer class="footer shell">
<div>
<strong>RinggAI</strong>
<p>Built by the RinggAI Team.</p>
</div>
<div class="footer-links">
<a href="https://ringg.ai/dashboard/stt" target="_blank" rel="noreferrer">Playground</a>
<a href="https://pypi.org/project/ringglabs/" target="_blank" rel="noreferrer">SDK</a>
<a href="https://huggingface.co/RinggAI" target="_blank" rel="noreferrer">Hugging Face</a>
</div>
</footer>
</body>
</html>
|