GPT-3.5 Turbo 16K
JSON →An extended-context variant of GPT-3.5 Turbo supporting up to 16,384 tokens for longer conversations and documents.
Specs
context window 16K tokens
max output 4K tokens
input price $3 / 1M tokens
output price $4 / 1M tokens
Capabilities
streamingfunction-callingjson-modetool-useprompt-caching
Dates
releasedJun 2023
knowledge cutoffSep 2021
deprecatedJul 2024
Resources
API
full doc /v1/models/gpt-3.5-turbo-16k