Skip to content

Commit d327b8e

Browse files
committed
Update README to clarify prompt caching provider support and remove redundant Ollama link
1 parent dca1562 commit d327b8e

1 file changed

Lines changed: 4 additions & 1 deletion

File tree

‎README.md‎

Lines changed: 4 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -74,11 +74,14 @@ Got tired with copy and paste? Autofill prompts with the browser extension:
7474
7575
## API tools
7676

77-
From code editing to voice input and commit messages, CWC supports your software development efforts on every step. Bring Your Own Key for a remote model provider of choice. Self-host on your local machine with [Ollama](https://ollama.com/search).
77+
From code editing to voice input and commit messages, CWC supports your software development efforts on every step. Bring Your Own Key for a remote model provider of choice. Self-host on your local machine.
7878

7979
> [!NOTE]
8080
> Thanks to smart context files ordering by update/pick recency, CWC heavily utilizes [prompt caching](https://developers.openai.com/api/docs/guides/prompt-caching/), across tasks. This consistently reduces input token costs by up to 90% and latency by up to 80%.
8181
82+
> [!IMPORTANT]
83+
> Not all providers equally support prompt caching. [Google](https://aistudio.google.com/api-keys) and [OpenAI](https://platform.openai.com/settings/organization/api-keys) are your best bets.
84+
8285
## Enterprise security
8386

8487
**Code Web Chat operates exclusively on your machine.** When using the browser extension for prompt autofill, data is moved via editor-browser communication channel run on local Websockets. For API tools, model providers are called directly.

0 commit comments

Comments
 (0)