Model Details
Released
May 7, 2026
Context Length
16,000
Max Output
2,000
Where is gpt-realtime-translate a perfect fit?
GPT-Realtime-Translate is OpenAI’s specialized realtime model for live speech translation, designed to translate spoken conversations while preserving natural conversational flow.
- Live multilingual conversations and voice applications
- Real-time interpretation for meetings, calls, and customer support
- Travel and hospitality applications requiring immediate spoken translation
- Contact centers serving customers across multiple languages
- Voice agents communicating with users in different languages
- International collaboration and remote meeting applications
- Applications where low translation latency matters more than general-purpose reasoning
- Speech translation workflows that need a more direct approach than chaining transcription, text translation, and TTS
- Live multilingual conversations and voice applications
- Real-time interpretation for meetings, calls, and customer support
- Travel and hospitality applications requiring immediate spoken translation
- Contact centers serving customers across multiple languages
- Voice agents communicating with users in different languages
- International collaboration and remote meeting applications
- Applications where low translation latency matters more than general-purpose reasoning
- Speech translation workflows that need a more direct approach than chaining transcription, text translation, and TTS
Quick Model Estimate
Your GPT-5 Cost Estimate
💰 Total Cost
USD 3.00
for 1000 input + 1000 output tokens
📥 Input (1000 × $N/A)
USD 1.5000
📤 Output (1000 × $0.034000)
USD 1.5000
Cost Breakdown
📥 Input 50%
📤 Output 50%
Prices updated daily from official provider data.
Pricing
|
Provider
↕
|
Modality
↕
|
Service tier
↕
|
Input Price
(per 1M tokens) ↕ |
Output Price
(per 1M tokens) ↕ |
Context Size
↕
|
View
|
|---|---|---|---|---|---|---|
 OpenAI | Audio | Standard | N/A | $0.0340 | 16,000 tokens | → |
Other Models in the gpt-realtime Family
FAQs about gpt-realtime-translate
What is gpt-realtime-translate's context window?
gpt-realtime-translate has a context window of 16,000 tokens, and returns up to 2,000 tokens in a single response.