Problem or Use Case
During setup for Hermes, the onboarding currently requires users to configure multiple API keys for different services. These include:
- Inference
- Tool calling
- Image generation
- Web search
- Other external capabilities
While inference can already work via the OpenAI compatible Venice API key, the rest of the functionality still requires separate keys. This creates friction during onboarding and breaks the privacy and simplicity narrative.
Proposed Solution
We're currently thinking that it may be a good idea to integrate Venice further into the onboarding flow. By doing so, users will only need to enter a single Venice API key during setup and it will reduce onboarding friction significantly.
Alternatives Considered
I guess technically the alternative is to just not do this or make it easier to use OpenAI compliant APIs in onboarding?
Feature Type
Configuration option
Scope
Medium (few files, < 300 lines)
Contribution
Problem or Use Case
During setup for Hermes, the onboarding currently requires users to configure multiple API keys for different services. These include:
While inference can already work via the OpenAI compatible Venice API key, the rest of the functionality still requires separate keys. This creates friction during onboarding and breaks the privacy and simplicity narrative.
Proposed Solution
We're currently thinking that it may be a good idea to integrate Venice further into the onboarding flow. By doing so, users will only need to enter a single Venice API key during setup and it will reduce onboarding friction significantly.
Alternatives Considered
I guess technically the alternative is to just not do this or make it easier to use OpenAI compliant APIs in onboarding?
Feature Type
Configuration option
Scope
Medium (few files, < 300 lines)
Contribution