Provide a collection of third-party APIs that give public access to AI capabilities.
WARNING: this crate is NOT PRODUCTION ready.
| Category | Variant | Tested Models | Stream |
| :--------------- | :------------- | :----------------------- | :----: |
| Chat completions | - | gpt-4, gpt-4-0314, gpt-4-0613, gpt-3.5-turbo, gpt-3.5-turbo-16k, gpt-3.5-turbo-16k-0613, gpt-3.5-turbo-0613, gpt-3.5-turbo-0301, text-davinci-003, text-davinci-002, code-davinci-002[^note_3] | yes |ext-davinci-002,code-davinci-002| yes |
| Completions | - |gpt-3.5-turbo,gpt-3.5-turbo-16k,gpt-3.5-turbo-16k-0613,gpt-3.5-turbo-0613,gpt-3.5-turbo-0301,text-davinci-003| yes | | yes |
| Edits | - |text-davinci-edit-001,code-davinci-edit-001| - |
| Images | Generations |dall-e[^note_1] | - |
| Images | Edits |dall-e[^note_1] | - |
| Images | Variations |dall-e[^note_1] | - |
| Embeddings | - |text-embedding-ada-002[^note_4] | - |
| Audios | Transcriptions |whisper-1| - |
| Audios | Translation |whisper-1| - |
| Moderation | - |text-moderation-latest[^note_2],text-moderation-stable` | - |
Note: - Updated on June 16th, 2023. - OpenAI's Fine Tunes endpoints are currently not supported.
| Category | Capability | Endpoint | | :---------------- | :--------- | :------------- | | Cognitive service | Speech | Text-to-Speech, Voice-List, Speech-to-Text (Batch) |
Note: - Azure CN is not supported by this repo yet.
Create a config.yml in your working directory, or root of this crate. Contents like this:
yaml
openai:
api_key: sk-XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
org_id: org-xxx
base_endpoint: https://api.openai.com
azure:
speech:
key: 4c7eXXXXXXXXXXXXXXXXXXXXXXX54c32
region: westus
If you prefer to load configuration from environment variables, set the contents to RUST_AI_CONFIG like example chat-completion-streamed does.
Currently, OpenAI doesn't support manually specify model for image related
tasks. So DALL·E models are not valid variants of Model enum.
text-moderation-004 model (March 23rd, 2023). But developers cannot use
its API endpoints to specify variants other than text-moderation-latest
and text-moderation-stable.
because developer currently only has access to 8K token ones (May 6th, 2023).
text-embedding-ada-002-v2 (March 23rd, 2023).