Summary
Add Requesty as a named LLM provider in the CLI configuration, mirroring the existing OpenRouter wiring.
Requesty is an OpenAI-compatible LLM gateway. It uses the same provider/model naming convention as OpenRouter (e.g. openai/gpt-4o-mini), so it fits the existing provider machinery without special handling.
Proposed change
In cli/nao_core/config/llm/__init__.py, alongside the existing OpenRouter entries:
LLMProvider enum: add REQUESTY = "requesty"PROVIDER_AUTH: ProviderAuthConfig(env_var="REQUESTY_API_KEY", api_key="required", base_url_env_var="REQUESTY_BASE_URL")DEFAULT_ANNOTATION_MODELS: LLMProvider.REQUESTY: "openai/gpt-4o-mini"promptConfig picker: questionary.Choice("Requesty", value="requesty")
This is a 6-line addition that mirrors OpenRouter 1:1.
Tested
- Verified live against
https://router.requesty.ai/v1/chat/completions with openai/gpt-4o-mini -> HTTP 200.
- Runtime-checked that
LLMProvider("requesty"), PROVIDER_AUTH, DEFAULT_ANNOTATION_MODELS, and LLMConfig(provider=REQUESTY, api_key=...) all resolve correctly.
I have a branch ready (add-requesty-provider) and opened PR #978, which was auto-closed by the contributor policy. Opening this issue per CONTRIBUTING.md — happy to have the PR reopened or to adjust scope/placement.
Disclosure: I work at Requesty.
Summary
Add Requesty as a named LLM provider in the CLI configuration, mirroring the existing OpenRouter wiring.
Requesty is an OpenAI-compatible LLM gateway. It uses the same
provider/modelnaming convention as OpenRouter (e.g.openai/gpt-4o-mini), so it fits the existing provider machinery without special handling.Proposed change
In
cli/nao_core/config/llm/__init__.py, alongside the existing OpenRouter entries:LLMProviderenum: addREQUESTY = "requesty"PROVIDER_AUTH:ProviderAuthConfig(env_var="REQUESTY_API_KEY", api_key="required", base_url_env_var="REQUESTY_BASE_URL")DEFAULT_ANNOTATION_MODELS:LLMProvider.REQUESTY: "openai/gpt-4o-mini"promptConfigpicker:questionary.Choice("Requesty", value="requesty")This is a 6-line addition that mirrors OpenRouter 1:1.
Tested
https://router.requesty.ai/v1/chat/completionswithopenai/gpt-4o-mini-> HTTP 200.LLMProvider("requesty"),PROVIDER_AUTH,DEFAULT_ANNOTATION_MODELS, andLLMConfig(provider=REQUESTY, api_key=...)all resolve correctly.I have a branch ready (
add-requesty-provider) and opened PR #978, which was auto-closed by the contributor policy. Opening this issue per CONTRIBUTING.md — happy to have the PR reopened or to adjust scope/placement.Disclosure: I work at Requesty.