LLM Reference

text-ada

Deprecated

About

OpenAI's text-embedding-ada-002 is a robust model adept at converting text into numerical vectors to encapsulate semantic meaning, making it ideal for tasks like semantic search, document similarity assessments, and text clustering. Though it has a knowledge cutoff at September 2021, limiting its awareness of newer events, it remains efficient and cost-effective. It may not excel in benchmarks involving classification, yet it proves beneficial for search engines, recommendation systems, and numerous natural language tasks. Despite the existence of newer models with enhanced multilingual capabilities, text-embedding-ada-002 is still favored for its balanced performance and affordability 2567.

Capabilities

MultimodalFunction CallingTool UseJSON Mode

API Versions

text-ada-001

Specifications

Released2020-06-01
Parameters350M
Context2K
ArchitectureDecoder Only
Knowledge cutoff2019-10
Specializationgeneral