gpt-realtime-2.1-mini

OpenAIgpt-realtime-2.1-mini

GPT-Realtime-2.1 Mini is a lightweight real-time multimodal model developed by OpenAI, optimized for low-latency voice conversations and interactive AI applications. It supports text, audio, and image inputs, generates natural speech and text responses, and provides reasoning, instruction-following, and tool-calling capabilities. Designed for faster responses and lower costs, it is well suited for voice assistants, intelligent customer service, conversational agents, and real-time multimodal experiences.

API Pricing

Price typeBilling expression
Official price
Base pricingPricing tier: base
  • Input usage × $5.00
  • Output usage × $10.00
  • Audio input × $10.00
  • Audio output × $20.00
GPT_Ultra×0.880% of official list price
Base pricingPricing tier: base
  • Input usage × $4.00
  • Output usage × $8.00
  • Audio input × $8.00
  • Audio output × $16.00

Group multiplier: x0.8

Official pricing is subject to the model provider's latest published pricing and may vary by request method or service region. GPT_Ultra — TokenByte channel pricing. Actual prices and availability reflect current platform data.

Model information

Model providerOpenAI
Model IDgpt-realtime-2.1-mini
Context window128000
InputTextImageAudio
OutputTextAudio

Frequently asked questions

01

What is GPT-Realtime-2.1 Mini?

GPT-Realtime-2.1 Mini is a distilled reasoning model from OpenAI, designed for fast, low-cost realtime voice interactions. It supports audio and text inputs through WebRTC, WebSocket, and SIP connections, with improved alphanumeric recognition compared to GPT-Realtime-2. It is particularly suited for realtime voice assistants, customer service automation, interactive voice applications, and intelligent voice agents.
02

How much does GPT-Realtime-2.1 Mini cost?

GPT-Realtime-2.1 Mini costs $0.60/M text input tokens and $2.40/M text output tokens, with separate rates for Cached Text Input at $0.06/M tokens, Audio Input at $10.00/M tokens, Cached Audio Input at $0.30/M tokens, Audio Output at $20.00/M tokens, Image Input at $0.80/M tokens, and Cached Image Input at $0.08/M tokens.
03

What is the context length of GPT-Realtime-2.1 Mini?

GPT-Realtime-2.1 Mini has a 128,000 token context window. It supports up to 32,000 output tokens.