openAI API token information
OpenAI v2 Embeddings - Part 2 (text-embedding-ada-002)
Number of tokens: 2675
| MODEL | VERSION | COST |
|---|---|---|
| Ada | v1 | $0.0107 |
| Babbage | v1 | $0.013375 |
| Curie | v1 | $0.0535 |
| Davinci | v1 | $0.535 |
| Ada | v2 | $0.00107 |
https://platform.openai.com/docs/guides/embeddings/what-are-embeddings
| MODEL | OUTPUT DIMENSIONS |
|---|---|
| Ada | 1024 |
| Babbage | 2048 |
| Curie | 4096 |
| Davinci | 12288 |
- 가장 많이 사용 문서 탐색
| USE CASES | AVAILABLE MODELS |
|---|---|
| Search, context relevance, information retrieval | text-search-ada-doc-001 text-search-ada-query-001 text-search-babbage-doc-001 text-search-babbage-query-001 text-search-curie-doc-001 text-search-curie-query-001 text-search-davinci-doc-001 text-search-davinci-query-001 |
-
Rate limits are enforced at the organization level, not user level, based on the specific endpoint used as well as the type of account you have.
-
Rate limits are measured in two ways: RPM (requests per minute) and TPM (tokens per minute).
-
TEXT & EMBEDDING
-
Free trial users •20 RPM •150,000 TPM
-
Pay-as-you-go users (first 48 hours) •60 RPM •250,000 TPM
-
Pay-as-you-go users (after 48 hours) •3,000 RPM •250,000 TPM
| TYPE | 1 TPM EQUALS |
|---|---|
| davinci | 1 token per minute |
| curie | 25 tokens per minute |
| babbage | 100 tokens per minute |
| ada | 200 tokens per minute |
test text / token_counts = 37
[kor_text]
금융통화위원회는 다음 통화정책방향 결정시까지 한국은행 기준금리를 현 수준(3.50%)에서 유지하여 통화정책을 운용하기로 하였다. 물가상승률이 둔화 흐름을지속하겠지만 상당기간 목표수준을 상회할 것으로 전망되는 만큼 현재의 긴축기조를 유지하는 것이 적절하다고 보았다. 추가 인상 필요성은 대내외 정책여건의 변화를 점검하면서 판단해 나갈 것이다..
[eng_text]
The old, creaky wooden door swung open slowly, revealing a dimly lit room filled with dusty books and antique furniture, while a faint scent of lavender lingered in the air, evoking a sense of nostalgia and mystery.