OpenAI has made GPT-Live-1.1 accessible in its API beneath the mannequin ID gpt-live-1.1. The GPT-Live-1.1 voice model is a full-duplex mannequin for multilingual conversations. It could possibly hear whereas it speaks. OpenAI staff states that it speaks with more natural and consistent accents across languages and completes duties extra reliably. The discharge additionally provides standby, a less expensive billing state for idle periods. For groups working voice brokers, idle time on a name now prices much less. You may test the deployment choices and specs right here by our GPULeaderboard.com chrome extension.
TL;DR
- Dimension: Reside periods maintain a 128,000-token default context window.
- Runs on: OpenAI’s cloud solely, by way of
v1/reside/periods. Shoppers join over WebRTC or WebSocket. No open weights, no quantized variants. - Efficiency: OpenAI claims higher accents, transcription, letter and quantity dealing with, and process completion than GPT-Reside-1. No scores revealed.
- Greatest: Standby payments idle WebRTC time at $0.02 per minute versus $0.05 active, a 60% decrease fee.
- Worst: OpenAI revealed 0 benchmark scores for 1.1. Picture, video and structured outputs are usually not supported.
- Backside line: Greatest: cheaper always-on voice periods. Worst: high quality features are self-reported with no numbers.
What’s GPT-Reside-1.1?
GPT-Reside-1.1 is a full-duplex speech mannequin from OpenAI that hears and speaks in 1 stream. It follows GPT-Live-1, which reached the API on September 10, 2026. As per the OpenAI forum announcement, it has powered ChatGPT Plus and Professional voice since October 15.
Inputs and outputs are audio and textual content. Picture and video are usually not supported. Streaming and performance calling work. Structured outputs, fine-tuning and predicted outputs don’t.
How does GPT-Reside-1.1 work?
The GPT-Reside household replaces the standard cascade of speech-to-text, an LLM and text-to-speech. OpenAI says the mannequin reasons over incoming and outgoing audio together in a single model. It additionally handles flip detection and returns ASR transcripts natively.
Deeper work is delegated. The voice mannequin can hand reasoning and power calls to a backend textual content mannequin. OpenAI names GPT-6 Astra for advanced points and Luna for high-volume duties. Third-party backends additionally work.
The Live conversations guide units delegation.sort to shopper or responses. Responses mode makes use of OpenAI fashions. Shopper mode connects your personal backend. Directions can run as much as 16,384 tokens.
What modified from GPT-Reside-1?
OpenAI lists 4 enchancment areas for 1.1:
- Extra pure and constant accents throughout languages
- Extra correct speech transcription
- Higher dealing with of exact particulars like letters and numbers
- Extra dependable instruction following and process completion
How does standby mode work?
Standby pauses inference throughout silence whereas the WebRTC connection stays open. The guide paperwork 3 settings:
session.wake_policy.sort: set tospeech, so inference waits for speechsession.idle_timeout_s: an entire quantity from 5 to 60 secondssession.wake_policy.vad_threshold: elective, 0 to 1, default 0.5
After the timeout, the server emits session.paused. When the person speaks, session.resumed fires. The session ID and historical past are stored. OpenAI warns the primary reply after a brief utterance could also be slower. Hold the microphone monitor lively throughout standby.
Standby works for each GPT-Reside-1 and GPT-Reside-1.1.
How a lot does GPT-Reside-1.1 price?
| Session state | Worth per minute | Worth per hour |
|---|---|---|
| Energetic | $0.05 | $3.00 |
| Standby (WebRTC, inference paused) | $0.02 | $1.20 |
Voice periods are billed by length. The GPT-Live-1 model page says billing is per second. Backend mannequin and power utilization are billed individually. That issues: a heavy backend can price greater than the voice layer.
Deployment and availability
The mannequin is API-only. You turn an current Reside session by setting mannequin to gpt-live-1.1. Altering the mannequin requires a brand new session.
Outbound SIP calling reached all organizations on November 2, so telephone brokers are supported. GPT-Reside-1 and GPT-Reside-1.1 have separate concurrent-session limits. Every restrict applies throughout inference areas. The 1.1 web page doesn’t publish tier numbers.
How does GPT-Reside-1.1 examine?
| Function | GPT-Reside-1.1 | GPT-Live-1 | GPT-Realtime-2.1 | Gemini 3.8 Live |
|---|---|---|---|---|
| Developer | OpenAI | OpenAI | OpenAI | |
| Parameters | Not disclosed | Not disclosed | Not disclosed | Not disclosed |
| Enter | Audio, textual content | Audio, textual content | Textual content, audio, picture | Textual content, picture, audio |
| Output | Audio, textual content | Audio, textual content | Textual content, audio | Textual content, audio |
| Context | 128K default (guide) | 128K default (guide) | 128K, 32K max output | 131,072 in, 65,536 out |
| Pricing | $0.05/min lively, $0.02/min standby | $0.05/min lively, $0.02/min standby | $32 / $64 per 1M audio tokens (in/out) | Token-based (pricing) |
| Idle standby billing | Sure (WebRTC) | Sure (WebRTC) | Not documented | Not documented |
| Reasoning | Delegated to backend mannequin | Delegated to backend mannequin | In-built, configurable effort | Interleaved reasoning |
| Perform calling | Sure | Sure | Sure | Sure |
| Structured outputs | No | No | Not disclosed | No |
| Printed benchmarks | None for 1.1 | Full Duplex Bench +30 pts vs GPT-Realtime-2.1 | Baseline in that eval | Not in contrast |
| License / weights | Closed, API solely | Closed, API solely | Closed, API solely | Closed, API solely |
#mtp-gl11{max-width:880px;margin:0 auto;background:#0d0d0d!necessary;coloration:#f2f2f2!necessary;border:1px strong #2a2a2a!necessary;border-radius:16px;font-family:Inter,system-ui,-apple-system,Segoe UI,Roboto,sans-serif;overflow:hidden;box-sizing:border-box}
#mtp-gl11 *{box-sizing:border-box}
#mtp-gl11 hr,#mtp-gl11 p:empty,#mtp-gl11 del,#mtp-gl11 s{show:none!necessary}
#mtp-gl11 .hd{padding:22px 24px 14px;background:radial-gradient(circle at 92% 0%,rgba(240,160,106,.25),clear 45%),radial-gradient(circle at 70% 120%,rgba(91,140,247,.22),clear 50%)!necessary}
#mtp-gl11 .tag{show:inline-block;font-size:11px;letter-spacing:.08em;text-transform:uppercase;coloration:#0d0d0d!necessary;background:#f2f2f2!necessary;border-radius:99px;padding:3px 10px;font-weight:700}
#mtp-gl11 h3{margin:10px 0 4px;font-size:22px;line-height:1.25;coloration:#fff!necessary}
#mtp-gl11 .sub{margin:0;coloration:#a9a9a9!necessary;font-size:14px}
#mtp-gl11 .tabs{show:flex;hole:6px;padding:0 24px;border-bottom:1px strong #262626!necessary;flex-wrap:wrap}
#mtp-gl11 .tab{background:clear!necessary;coloration:#9a9a9a!necessary;border:0!necessary;border-bottom:2px strong clear!necessary;padding:12px 10px;font:600 14px Inter,system-ui,sans-serif;cursor:pointer}
#mtp-gl11 .tab.on{coloration:#fff!necessary;border-bottom-color:#F0A06A!necessary}
#mtp-gl11 .pn{show:none;padding:20px 24px 8px;animation:gl11fade .35s ease}
#mtp-gl11 .pn.on{show:block}
@keyframes gl11fade{from{opacity:0;rework:translateY(6px)}to{opacity:1;rework:none}}
#mtp-gl11 .row{show:flex;hole:10px;align-items:middle;flex-wrap:wrap}
#mtp-gl11 .btn{background:#f2f2f2!necessary;coloration:#0d0d0d!necessary;border:0!necessary;border-radius:99px;padding:8px 16px;font:700 13px Inter,system-ui,sans-serif;cursor:pointer}
#mtp-gl11 .btn.gh{background:clear!necessary;coloration:#f2f2f2!necessary;border:1px strong #3a3a3a!necessary}
#mtp-gl11 .notice{font-size:12px;coloration:#8a8a8a!necessary;margin:10px 0}
#mtp-gl11 .field{background:#151515!necessary;border:1px strong #2a2a2a!necessary;border-radius:12px;padding:14px}
#mtp-gl11 svg textual content{font-family:Inter,system-ui,sans-serif}
#mtp-gl11 .node{transition:all .4s}
#mtp-gl11 .movement{stroke-dasharray:6 6;animation:gl11dash 1s linear infinite}
@keyframes gl11dash{to{stroke-dashoffset:-12}}
#mtp-gl11 .pulse{animation:gl11pulse 1.6s ease-in-out infinite;transform-origin:middle;transform-box:fill-box}
@keyframes gl11pulse{0%,100%{rework:scale(1);opacity:1}50%{rework:scale(1.08);opacity:.85}}
#mtp-gl11 .dim{opacity:.25}
#mtp-gl11 .stepTxt{min-height:44px;font-size:14px;coloration:#ddd!necessary;margin:8px 0}
#mtp-gl11 .tl{show:flex;peak:34px;border-radius:8px;overflow:hidden;border:1px strong #2a2a2a!necessary;margin:12px 0 6px}
#mtp-gl11 .seg{show:flex;align-items:middle;justify-content:middle;font-size:11px;font-weight:700;coloration:#0d0d0d!necessary;transition:flex-grow .6s,opacity .4s;opacity:.2;white-space:nowrap;overflow:hidden}
#mtp-gl11 .seg.lit{opacity:1}
#mtp-gl11 .log{font:12px/1.6 ui-monospace,Menlo,Consolas,monospace;background:#0a0a0a!necessary;coloration:#cfcfcf!necessary;border:1px strong #232323!necessary;border-radius:8px;padding:10px;min-height:96px;margin:0;white-space:pre-wrap;word-break:break-word}
#mtp-gl11 .log b{coloration:#F0A06A!necessary}
#mtp-gl11 enter[type=range]{width:100%;accent-color:#F0A06A}
#mtp-gl11 label{font-size:13px;coloration:#bbb!necessary;show:block;margin:10px 0 4px}
#mtp-gl11 .val{coloration:#fff!necessary;font-weight:700}
#mtp-gl11 .bars{margin-top:14px}
#mtp-gl11 .bar{margin:8px 0}
#mtp-gl11 .bar .lb{show:flex;justify-content:space-between;font-size:13px;margin-bottom:4px;coloration:#ddd!necessary}
#mtp-gl11 .bar .tr{peak:16px;background:#1c1c1c!necessary;border-radius:99px;overflow:hidden}
#mtp-gl11 .bar .fl{peak:100%;width:0;border-radius:99px;transition:width .8s cubic-bezier(.2,.8,.2,1)}
#mtp-gl11 .grid{show:grid;grid-template-columns:1fr 1fr;hole:10px;align-items:begin}
#mtp-gl11 .card{background:#151515!necessary;border:1px strong #2a2a2a!necessary;border-radius:12px;padding:12px 14px;cursor:pointer;transition:border-color .25s,rework .25s}
#mtp-gl11 .card:hover{border-color:#5B8CF7!necessary;rework:translateY(-2px)}
#mtp-gl11 .card h4{margin:0;font-size:14px;coloration:#fff!necessary}
#mtp-gl11 .card p{margin:6px 0 0;font-size:13px;coloration:#aaa!necessary;max-height:0;overflow:hidden;transition:max-height .4s}
#mtp-gl11 .card.open p{max-height:120px}
#mtp-gl11 .card .plus{float:proper;coloration:#F0A06A!necessary;font-weight:700}
#mtp-gl11 code{background:#1f1f1f!necessary;coloration:#F0A06A!necessary;border:0!necessary;padding:1px 5px;border-radius:4px;font-size:12px}
#mtp-gl11 .ft{show:flex;justify-content:space-between;hole:10px;flex-wrap:wrap;padding:12px 24px 16px;border-top:1px strong #222!necessary;font-size:12px;coloration:#777!necessary;margin-top:12px}
#mtp-gl11 .ft a{coloration:#9a9a9a!necessary}
#mtp-gl11 .ft .model{coloration:#76B900!necessary;font-weight:700}
@media (max-width:640px){#mtp-gl11 .hd{padding:18px 16px 12px}#mtp-gl11 h3{font-size:19px}#mtp-gl11 .tabs{padding:0 12px}#mtp-gl11 .tab{font-size:12px;padding:10px 6px}#mtp-gl11 .pn{padding:16px}#mtp-gl11 .grid{grid-template-columns:1fr}#mtp-gl11 .seg{font-size:9px}#mtp-gl11 .ft{padding:12px 16px}}
</fashion>
<div id="mtp-gl11">
<div class="hd">
<span class="tag">OpenAI · gpt-live-1.1</span>
<h3>GPT-Reside-1.1, defined</h3>
<p class="sub">Full-duplex voice mannequin within the API. $0.05/min lively, $0.02/min standby.</p>
</div>
<div class="tabs" position="tablist">
<button class="tab on" data-t="0">The way it works</button>
<button class="tab" data-t="1">Standby mode</button>
<button class="tab" data-t="2">Value calculator</button>
<button class="tab" data-t="3">What modified</button>
</div>
<div class="pn on" id="gl11p0">
<svg viewBox="0 0 820 220" width="100%" position="img" aria-label="GPT-Reside-1.1 structure">
<defs>
<radialGradient id="gl11orb" cx="35%" cy="35%"><cease offset="0" stop-color="#FBD3B0"/><cease offset=".6" stop-color="#F0A06A"/><cease offset="1" stop-color="#D9774A"/></radialGradient>
<radialGradient id="gl11orb2" cx="35%" cy="35%"><cease offset="0" stop-color="#D6E2FF"/><cease offset=".6" stop-color="#7EA4FA"/><cease offset="1" stop-color="#3F74F2"/></radialGradient>
</defs>
<g id="gl11n0" class="node">
<rect x="10" y="70" width="160" peak="80" rx="14" fill="#151515" stroke="#3a3a3a"/>
<textual content x="90" y="104" text-anchor="center" fill="#fff" font-size="15" font-weight="700">Consumer audio</textual content>
<textual content x="90" y="126" text-anchor="center" fill="#999" font-size="12">WebRTC / WebSocket / SIP</textual content>
</g>
<path id="gl11a0" class="movement" d="M172 100 H300" stroke="#F0A06A" stroke-width="3" fill="none"/>
<path id="gl11a1" class="movement" d="M300 122 H172" stroke="#7EA4FA" stroke-width="3" fill="none"/>
<g id="gl11n1" class="node">
<circle class="pulse" cx="380" cy="110" r="62" fill="url(#gl11orb)"/>
<textual content x="380" y="106" text-anchor="center" fill="#1a0d05" font-size="15" font-weight="800">GPT-Reside-1.1</textual content>
<textual content x="380" y="126" text-anchor="center" fill="#3a1d0b" font-size="11" font-weight="600">listens whereas talking</textual content>
</g>
<path id="gl11a2" class="movement" d="M444 96 H632" stroke="#7EA4FA" stroke-width="3" fill="none"/>
<path id="gl11a3" class="movement" d="M632 126 H444" stroke="#F0A06A" stroke-width="3" fill="none"/>
<g id="gl11n2" class="node">
<circle cx="700" cy="110" r="66" fill="url(#gl11orb2)"/>
<textual content x="700" y="104" text-anchor="center" fill="#071436" font-size="14" font-weight="800">Backend mannequin</textual content>
<textual content x="700" y="122" text-anchor="center" fill="#0d2260" font-size="10" font-weight="600">Astra · Luna · third get together</textual content>
</g>
<textual content x="236" y="88" text-anchor="center" fill="#bbb" font-size="11">speech in</textual content>
<textual content x="236" y="142" text-anchor="center" fill="#bbb" font-size="11">speech out</textual content>
<textual content x="522" y="84" text-anchor="center" fill="#bbb" font-size="11">delegate instruments</textual content>
<textual content x="522" y="146" text-anchor="center" fill="#bbb" font-size="11">consequence</textual content>
</svg>
<div class="stepTxt" id="gl11st">One mannequin handles listening and talking. No separate speech-to-text and text-to-speech chain.</div>
<div class="row">
<button class="btn" id="gl11next">Step by ▸</button>
<button class="btn gh" id="gl11all">Present all</button>
<span class="notice" id="gl11ct" fashion="margin:0">Overview</span>
</div>
<p class="notice">Delegation modes: <code>delegation.sort</code> = <code>shopper</code> (your backend) or <code>responses</code> (OpenAI fashions). Default context: 128,000 tokens.</p>
</div>
<div class="pn" id="gl11p1">
<label>Idle timeout <code>session.idle_timeout_s</code>: <span class="val" id="gl11tv">20</span> s (allowed 5 to 60)</label>
<enter sort="vary" id="gl11to" min="5" max="60" worth="20">
<div class="tl">
<div class="seg" id="gl11s0" fashion="flex-grow:3;background:#F0A06A">Energetic · speaking</div>
<div class="seg" id="gl11s1" fashion="flex-grow:2;background:#d9c2b0">Silence (timeout)</div>
<div class="seg" id="gl11s2" fashion="flex-grow:4;background:#7EA4FA">Standby · $0.02/min</div>
<div class="seg" id="gl11s3" fashion="flex-grow:2;background:#F0A06A">Resumed</div>
</div>
<div class="row" fashion="margin:8px 0 10px"><button class="btn" id="gl11sim"> Simulate a name</button><button class="btn gh" id="gl11rst">Reset</button></div>
<pre class="log" id="gl11log">Press Simulate to stroll by the standby occasions.</pre>
<p class="notice">WebRTC solely. Hold the mic monitor lively. Session ID and historical past are stored. First reply after a brief utterance could also be slower.</p>
</div>
<div class="pn" id="gl11p2">
<label>Session size: <span class="val" id="gl11mv">60</span> min</label>
<enter sort="vary" id="gl11min" min="1" max="480" worth="60">
<label>Share of the session that’s idle (in standby): <span class="val" id="gl11iv">50</span>%</label>
<enter sort="vary" id="gl11idle" min="0" max="95" worth="50">
<div class="bars">
<div class="bar"><div class="lb"><span>All minutes at lively fee ($0.05)</span><span class="val" id="gl11c0">$3.00</span></div><div class="tr"><div class="fl" id="gl11b0" fashion="background:#F0A06A"></div></div></div>
<div class="bar"><div class="lb"><span>With standby ($0.02 on idle minutes)</span><span class="val" id="gl11c1">$2.10</span></div><div class="tr"><div class="fl" id="gl11b1" fashion="background:#7EA4FA"></div></div></div>
</div>
<p class="notice" id="gl11sv">Voice layer solely. Illustrative: the timeout window itself payments on the lively fee, and backend mannequin and power utilization are billed individually.</p>
</div>
<div class="pn" id="gl11p3">
<div class="grid">
<div class="card"><h4>Accents <span class="plus">+</span></h4><p>Extra pure and constant accents throughout languages, per OpenAI.</p></div>
<div class="card"><h4>Transcription <span class="plus">+</span></h4><p>Transcribes speech extra precisely than GPT-Reside-1, per the mannequin web page.</p></div>
<div class="card"><h4>Letters and numbers <span class="plus">+</span></h4><p>Higher dealing with of exact particulars like letters and numbers.</p></div>
<div class="card"><h4>Job completion <span class="plus">+</span></h4><p>Extra dependable instruction following and process completion.</p></div>
</div>
<div class="bars">
<div class="bar"><div class="lb"><span>Energetic fee</span><span class="val">$0.05/min</span></div><div class="tr"><div class="fl gl11pb" data-w="100" fashion="background:#F0A06A"></div></div></div>
<div class="bar"><div class="lb"><span>Standby fee (new, additionally on GPT-Reside-1)</span><span class="val">$0.02/min</span></div><div class="tr"><div class="fl gl11pb" data-w="40" fashion="background:#7EA4FA"></div></div></div>
</div>
<p class="notice">OpenAI revealed no benchmark scores for 1.1. All 4 features above are self-reported.</p>
</div>
<div class="ft">
<span>Sources: <a href="https://builders.openai.com/api/docs/fashions/gpt-live-1.1" goal="_blank" rel="noopener">mannequin web page</a> · <a href="https://builders.openai.com/api/docs/guides/live-conversations" goal="_blank" rel="noopener">Reside information</a></span>
<span class="model">Constructed by Marktechpost</span>
</div>
</div>
<script>
(operate(){
var R=doc.getElementById(‘mtp-gl11′);
operate $(i){return doc.getElementById(i)}
operate resize(){strive{father or mother.postMessage({mtpFrame:’gl11′,peak:R.offsetHeight+40},’*’)}catch(e){}}
var tabs=R.querySelectorAll(‘.tab’);
for(var i=0;i<tabs.size;i++){tabs[i].addEventListener(‘click on’,operate(){
for(var j=0;j<tabs.size;j++){tabs[j].classList.take away(‘on’);$(‘gl11p’+j).classList.take away(‘on’)}
this.classList.add(‘on’);var t=this.getAttribute(‘data-t’);$(‘gl11p’+t).classList.add(‘on’);
if(t===’2′)calc(); if(t===’3′)priceBars(); setTimeout(resize,60);
})}
/* step-through */
var steps=[
{on:[‘gl11n0′,’gl11a0′,’gl11n1′],t:’1. Caller audio streams in. The mannequin retains listening even whereas it’s talking (full duplex).’,c:’Step 1 of three’},
{on:[‘gl11n1′,’gl11a2′,’gl11n2′],t:’2. For reasoning or instruments, the voice mannequin delegates to a backend mannequin whereas the dialog continues.’,c:’Step 2 of three’},
{on:[‘gl11n2′,’gl11a3′,’gl11n1′,’gl11a1′,’gl11n0′],t:’3. The consequence returns and is spoken again. ASR transcripts and response textual content come again natively.’,c:’Step 3 of three’}];
var ids=[‘gl11n0′,’gl11n1′,’gl11n2′,’gl11a0′,’gl11a1′,’gl11a2′,’gl11a3’],si=-1;
operate showAll(){ids.forEach(operate(x){$(x).classList.take away(‘dim’)})}
$(‘gl11next’).addEventListener(‘click on’,operate(){si=(si+1)%steps.size;var s=steps[si];
ids.forEach(operate(x){$(x).classList.toggle(‘dim’,s.on.indexOf(x)<0)});$(‘gl11st’).textContent=s.t;$(‘gl11ct’).textContent=s.c;});
$(‘gl11all’).addEventListener(‘click on’,operate(){si=-1;showAll();$(‘gl11st’).textContent=”One mannequin handles listening and talking. No separate speech-to-text and text-to-speech chain.”;$(‘gl11ct’).textContent=”Overview”});
/* standby */
$(‘gl11to’).addEventListener(‘enter’,operate(){$(‘gl11tv’).textContent=this.worth;$(‘gl11s1’).fashion.flexGrow=(1+this.worth/15).toFixed(2)});
var timers=[];
operate rst(){timers.forEach(clearTimeout);timers=[];for(var ok=0;ok<4;ok++)$(‘gl11s’+ok).classList.take away(‘lit’);$(‘gl11log’).innerHTML=’Press Simulate to stroll by the standby occasions.’}
$(‘gl11rst’).addEventListener(‘click on’,rst);
$(‘gl11sim’).addEventListener(‘click on’,operate(){rst();var to=$(‘gl11to’).worth;var L=$(‘gl11log’);L.innerHTML=”;
var ev=[[0,'<b>session.resumed</b> first wake on speech, inference active ($0.05/min)’],
[1,’silence detected, waiting ‘+to+’ s (idle_timeout_s)’],
[2,'<b>session.paused</b> standby, WebRTC call stays open ($0.02/min)’],
[3,’user speaks, <b>session.resumed</b> same session ID, history restored’]];
ev.forEach(operate(e,n){timers.push(setTimeout(operate(){$(‘gl11s’+e[0]).classList.add(‘lit’);L.innerHTML+=(n?’n’:”)+e[1];resize()},n*1100))});
});
/* calculator */
operate cash(v){return ‘$’+v.toFixed(2)}
operate calc(){var m=+$(‘gl11min’).worth,p=+$(‘gl11idle’).worth/100;$(‘gl11mv’).textContent=m;$(‘gl11iv’).textContent=Math.spherical(p*100);
var a=m*0.05,b=m*(1-p)*0.05+m*p*0.02;$(‘gl11c0’).textContent=cash(a);$(‘gl11c1’).textContent=cash(b);
$(‘gl11b0’).fashion.width=”100%”;$(‘gl11b1′).fashion.width=(a?b/a*100:0)+’%’;
$(‘gl11sv’).textContent=”Saves “+cash(a-b)+’ (‘+(a?((a-b)/a*100).toFixed(0):0)+’%) on the voice layer. Illustrative: the timeout window itself payments on the lively fee, and backend mannequin and power utilization are billed individually.’}
$(‘gl11min’).addEventListener(‘enter’,calc);$(‘gl11idle’).addEventListener(‘enter’,calc);
operate priceBars(){var b=R.querySelectorAll(‘.gl11pb’);for(var ok=0;ok<b.size;ok++){b[k].fashion.width=”0″;(operate(el){setTimeout(operate(){el.fashion.width=el.getAttribute(‘data-w’)+’%’},50)})(b[k])}}
var playing cards=R.querySelectorAll(‘.card’);for(var c=0;c<playing cards.size;c++){playing cards[c].addEventListener(‘click on’,operate(){this.classList.toggle(‘open’);this.querySelector(‘.plus’).textContent=this.classList.incorporates(‘open’)?’−’:’+’;setTimeout(resize,420)})}
window.addEventListener(‘load’,resize);window.addEventListener(‘resize’,resize);setTimeout(resize,300);
})();
</script>
</physique></html>”>
window.addEventListener(“message”,operate(e){if(e.knowledge&&e.knowledge.mtpFrame===”gl11″&&e.knowledge.peak){doc.getElementById(“mtp-gl11-frame”).fashion.peak=e.knowledge.peak+”px”;}});
Key Takeaways
- GPT-Reside-1.1 is reside within the API as
gpt-live-1.1. - Standby cuts idle WebRTC billing from $0.05 to $0.02 per minute.
- Idle timeout is configurable from 5 to 60 seconds.
- Accent, transcription and process features are self-reported, with no scores.
- Backend reasoning and instruments are billed individually from voice minutes.
Take a look at the model page, the Live conversations guide and the announcement. You may test the deployment choices and specs right here by our GPULeaderboard.com chrome extension.
Additionally, be at liberty to observe us on Twitter and don’t neglect to hitch our 150k+ML SubReddit and Subscribe to our Publication. Wait! are you on telegram? now you can join us on telegram as well.
Must associate with us for selling your GitHub Repo OR Hugging Face Web page OR Product Launch OR Webinar and so forth.? Connect with us
The put up OpenAI Ships GPT-Reside-1.1 for Voice Brokers, Cuts Idle Value to $0.02/Min appeared first on MarkTechPost.

