GPT Realtime
JSON →A dated snapshot of the GPT Realtime model, pinned for stability in production deployments.
Specs
context window 32K tokens
max output 4K tokens
input price $4 / 1M tokens
output price $16 / 1M tokens
Capabilities
streamingtool-usefunction-calling
Dates
releasedAug 2025
knowledge cutoffJun 2025
Aliases
gpt-realtime Resources
API
full doc /v1/models/gpt-realtime-2025-08-28