Feature Description
Add Modelsell as a named OpenAI-compatible model provider integration in Flowise.
Modelsell provides a unified OpenAI-compatible endpoint:
The integration should reuse Flowise's existing OpenAI-compatible model path and fetch models dynamically from /v1/models instead of maintaining a hardcoded model list.
Feature Category
Integration
Problem Statement
Modelsell can be configured manually through an OpenAI-compatible/custom model node today, but users need to know and repeatedly enter the correct endpoint, key location, and model IDs. This makes the provider harder to discover and introduces avoidable configuration errors across flows.
Proposed Solution
Add a focused Modelsell provider preset or model component that:
- Defaults the base URL to
https://modelsell.com/v1.
- Uses bearer API-key authentication.
- Loads available models from
GET /v1/models.
- Reuses the existing OpenAI Chat Completions implementation without provider-specific request transformations.
- Includes links to the Modelsell API key page and documentation.
If a dedicated component is not preferred for OpenAI-compatible providers, a documented preset/example using the existing custom OpenAI component would still reduce setup friction. I can contribute the implementation in the form preferred by the maintainers.
Mockups or References
Additional Context
The first contribution can stay limited to Chat Completions and dynamic model discovery. It does not need a hardcoded catalog or a new transport layer.
Disclosure: I represent Modelsell.
Feature Description
Add Modelsell as a named OpenAI-compatible model provider integration in Flowise.
Modelsell provides a unified OpenAI-compatible endpoint:
https://modelsell.com/v1POST /v1/chat/completionsGET /v1/modelsThe integration should reuse Flowise's existing OpenAI-compatible model path and fetch models dynamically from
/v1/modelsinstead of maintaining a hardcoded model list.Feature Category
Integration
Problem Statement
Modelsell can be configured manually through an OpenAI-compatible/custom model node today, but users need to know and repeatedly enter the correct endpoint, key location, and model IDs. This makes the provider harder to discover and introduces avoidable configuration errors across flows.
Proposed Solution
Add a focused Modelsell provider preset or model component that:
https://modelsell.com/v1.GET /v1/models.If a dedicated component is not preferred for OpenAI-compatible providers, a documented preset/example using the existing custom OpenAI component would still reduce setup friction. I can contribute the implementation in the form preferred by the maintainers.
Mockups or References
Additional Context
The first contribution can stay limited to Chat Completions and dynamic model discovery. It does not need a hardcoded catalog or a new transport layer.
Disclosure: I represent Modelsell.