What are the official context and output token limits for gpt-audio-1.5 API?

gpt-audio-1.5 supports a 128,000-token context window and documents list a max output token configuration around 16,384; verify exact limits per endpoint in the developer docs. :contentReference[oaicite:44]{index=44}

Can gpt-audio-1.5 handle both speech-to-text and text-to-speech in the API?

Yes — it accepts audio inputs and can return audio outputs or textual responses via the Chat Completions/audio endpoints. :contentReference[oaicite:45]{index=45}

When should I use gpt-audio-1.5 vs gpt-realtime-1.5 for a voice agent?

Choose gpt-audio-1.5 for higher-quality audio in Chat Completions flows where larger context is required; choose gpt-realtime-1.5 for low-latency, live streaming voice interactions. :contentReference[oaicite:46]{index=46}

Does gpt-audio-1.5 support streaming and function calling for tool integrations?

Yes — the model supports streaming audio responses and structured outputs/function calling to integrate external tools and workflows. :contentReference[oaicite:47]{index=47}

Is gpt-audio-1.5 suitable for production customer support voice agents?

Yes — it's designed for voice assistants and conversational agents, but you should add human review/QA, logging, and safety controls before production deployment. :contentReference[oaicite:48]{index=48}

What are the main limitations to consider when deploying gpt-audio-1.5?

Key considerations are compute/latency tradeoffs for large context audio sessions, safety guardrails for voice content, and the need to validate ASR/TTS outputs in your domain. :contentReference[oaicite:49]{index=49}

API gpt-audio-1.5 Acessível | text-to-speech

Technical specifications of gpt-audio-1.5

Item	gpt-audio-1.5 (public specs)
Model family	GPT Audio family (audio-first variant)
Input types	Text, audio (speech in)
Output types	Text, audio (speech out), structured outputs (function calls supported)
Context window	128,000 tokens.
Max output tokens	16,384 (documented in related gpt-audio listing).
Performance tier	Higher intelligence; Medium speed (balanced).
Latency profile	Optimized for voice interactions (mid/low latency depending on endpoint).
Availability	Chat Completions API (audio in/out) and platform playgrounds; integrated across realtime/voice surfaces.
Safety / usage notes	Guardrails for voice content; treat model outputs with the usual safety and verification for production voice agents.

Note: gpt-realtime-1.5 is a closely related realtime audio/voice-first variant optimized for lower latency and realtime sessions; compare below.

What is gpt-audio-1.5?

gpt-audio-1.5 is an audio-capable GPT model that supports both speech input and speech output through the Chat Completions and related audio-capable APIs. It's positioned as the main generally-available audio model for building voice agents and speech-first experiences while balancing quality and speed.

Main features

Speech-in / speech-out support: Handle spoken input and return spoken or textual responses for natural voice flows.
Large context for audio workflows: Supports very large context (documented 128k tokens) enabling multi-turn, long conversation history or large multimodal sessions.
Streaming & Chat Completions compatibility: Works inside Chat Completions with streaming audio responses and function-call structured outputs.
Balanced performance/latency: Tuned to provide high quality audio responses at medium throughput—suitable for chatbots and voice assistants where quality matters.
Ecosystem & integrations: Supported in the platform’s playgrounds and available across official realtime/voice endpoints and partner integrations (Azure/Microsoft Foundry notes reference similar audio models).

Property	gpt-audio-1.5	gpt-realtime-1.5
Primary focus	High-quality audio in/out for Chat Completions and conversational flows.	Realtime S2S (speech-to-speech) with lower latency for live voice agents and streaming scenarios.
Context window	128k tokens.	32k tokens (realtime variant documented).
Max output tokens	16,384 (documented).	Typically configured for shorter realtime responses (docs list smaller max tokens).
Best use	Chatbots, voice-enabled assistants where full chat semantics + audio are required.	Live voice agents, kiosks, and low-latency conversational interfaces.

Representative use cases

Conversational voice agents for customer support and internal help desks.
Voice-enabled assistants embedded in apps, devices, and kiosks.
Hands-free workflows (dictation, voice search, accessibility).
Multimodal experiences that mix audio with text / images via Chat Completions.

Limitations & operational considerations

Not a drop-in replacement for human QA: Always validate speech outputs and downstream actions with human review in production flows.
Resource planning: Large context and audio I/O can increase compute and latency—design streaming/segmentation strategies for long sessions.
Safety & policy constraints: Voice outputs can carry persuasive power; follow platform safety guidelines and guardrails when deploying at scale.
How to access GPT Audio 1.5 API

Log in to cometapi.com. If you are not our user yet, please register first. Sign into your CometAPI console. Get the access credential API key of the interface. Click “Add Token” at the API token in the personal center, get the token key: sk-xxxxx and submit.

cometapi-key

Step 2: Send Requests to GPT Audio 1.5 API

Select the “gpt-audio-1.5” endpoint to send the API request and set the request body. The request method and request body are obtained from our website API doc. Our website also provides Apifox test for your convenience. Replace <YOUR_API_KEY> with your actual CometAPI key from your account. base url is Chat Completions

Insert your question or request into the content field—this is what the model will respond to . Process the API response to get the generated answer.

Step 3: Retrieve and Verify Results

Process the API response to get the generated answer. After processing, the API responds with the task status and output data.

Preços para gpt-audio-1.5

Explore preços competitivos para gpt-audio-1.5, projetado para atender diversos orçamentos e necessidades de uso. Nossos planos flexíveis garantem que você pague apenas pelo que usar, facilitando o dimensionamento conforme suas necessidades crescem. Descubra como gpt-audio-1.5 pode aprimorar seus projetos mantendo os custos gerenciáveis.

Comet Price (USD / M Tokens)	Official Price (USD / M Tokens)	Discount
Entrada:$2/M Saída:$8/M	Entrada:$2.5/M Saída:$10/M	-20%

Código de exemplo e API para gpt-audio-1.5

Acesse código de exemplo abrangente e recursos de API para gpt-audio-1.5 para otimizar seu processo de integração. Nossa documentação detalhada fornece orientação passo a passo, ajudando você a aproveitar todo o potencial do gpt-audio-1.5 em seus projetos.

# Get your CometAPI key from https://api.cometapi.com/console/token
# Export it as: export COMETAPI_KEY="your-key-here"

RESPONSE=$(curl https://api.cometapi.com/v1/chat/completions \
  -s \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $COMETAPI_KEY" \
  -d '{
    "model": "gpt-audio-1.5",
    "modalities": ["text", "audio"],
    "audio": {
      "voice": "alloy",
      "format": "wav"
    },
    "messages": [
      {
        "role": "user",
        "content": "Is a golden retriever a good family dog?"
      }
    ]
  }')

# Print the text transcript
echo "$RESPONSE" | python3 -c "import sys, json; r=json.load(sys.stdin); print(r['choices'][0]['message']['audio']['transcript'])"

# Save the audio to a WAV file
echo "$RESPONSE" | python3 -c "
import sys, json, base64
r = json.load(sys.stdin)
audio_data = r['choices'][0]['message']['audio']['data']
with open('gpt-audio-1.5-output.wav', 'wb') as f:
    f.write(base64.b64decode(audio_data))
print('Audio saved to gpt-audio-1.5-output.wav')
"

cURL Code Example

# Get your CometAPI key from https://api.cometapi.com/console/token
# Export it as: export COMETAPI_KEY="your-key-here"

RESPONSE=$(curl https://api.cometapi.com/v1/chat/completions \
  -s \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $COMETAPI_KEY" \
  -d '{
    "model": "gpt-audio-1.5",
    "modalities": ["text", "audio"],
    "audio": {
      "voice": "alloy",
      "format": "wav"
    },
    "messages": [
      {
        "role": "user",
        "content": "Is a golden retriever a good family dog?"
      }
    ]
  }')

# Print the text transcript
echo "$RESPONSE" | python3 -c "import sys, json; r=json.load(sys.stdin); print(r['choices'][0]['message']['audio']['transcript'])"

# Save the audio to a WAV file
echo "$RESPONSE" | python3 -c "
import sys, json, base64
r = json.load(sys.stdin)
audio_data = r['choices'][0]['message']['audio']['data']
with open('gpt-audio-1.5-output.wav', 'wb') as f:
    f.write(base64.b64decode(audio_data))
print('Audio saved to gpt-audio-1.5-output.wav')
"

Python Code Example

from openai import OpenAI
import os
import base64

# Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here
COMETAPI_KEY = os.environ.get("COMETAPI_KEY") or "<YOUR_COMETAPI_KEY>"
BASE_URL = "https://api.cometapi.com/v1"

client = OpenAI(base_url=BASE_URL, api_key=COMETAPI_KEY)

completion = client.chat.completions.create(
    model="gpt-audio-1.5",
    modalities=["text", "audio"],
    audio={"voice": "alloy", "format": "wav"},
    messages=[
        {
            "role": "user",
            "content": "Is a golden retriever a good family dog?"
        }
    ],
)

# Print the text response
print(completion.choices[0].message.audio.transcript)

# Save the audio response to a file
wav_bytes = base64.b64decode(completion.choices[0].message.audio.data)
output_path = "gpt-audio-1.5-output.wav"
with open(output_path, "wb") as f:
    f.write(wav_bytes)
print(f"Audio saved to {output_path}")

JavaScript Code Example

import OpenAI from "openai";
import fs from "fs";

// Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here
const api_key = process.env.COMETAPI_KEY || "<YOUR_COMETAPI_KEY>";
const base_url = "https://api.cometapi.com/v1";

const openai = new OpenAI({ apiKey: api_key, baseURL: base_url });

const completion = await openai.chat.completions.create({
  model: "gpt-audio-1.5",
  modalities: ["text", "audio"],
  audio: { voice: "alloy", format: "wav" },
  messages: [
    {
      role: "user",
      content: "Is a golden retriever a good family dog?",
    },
  ],
});

// Print the text transcript
console.log(completion.choices[0].message.audio.transcript);

// Save the audio response to a file
const wavBytes = Buffer.from(completion.choices[0].message.audio.data, "base64");
const outputPath = "gpt-audio-1.5-output.wav";
fs.writeFileSync(outputPath, wavBytes);
console.log(`Audio saved to ${outputPath}`);

gpt-audio-1.5