Function getInferenceClient

src/services/api/inferenceClient.ts:86–127

(
  args: Parameters<typeof getAnthropicClient>[0],
)

}

export async function getInferenceClient(
  args: Parameters<typeof getAnthropicClient>[0],
): Promise<InferenceClient> {
  if (isOpenAICompatByokActive()) {
    const apiKey = getDirectApiKeyEnvValue()
    const baseURL = getOpenAICompatBaseUrl()
    if (apiKey && baseURL) {
      return new OpenAICompatInferenceClient({
        baseURL,
        headers: getOpenAICompatByokHeaders(apiKey),
        useNCodeManagedModelRouting: false,
        wsV2Transport: null,
        ...(args.fetchOverride ? { fetch: args.fetchOverride } : {}),
      })
    }
  }

  if (getAPIProvider() === 'firstParty') {
    const managedModelBaseURL = getNCodeManagedModelBaseUrl(args.model)
    const configuredCompatBaseURL =
      getNoumenaBaseUrl() ?? getLegacyOpenAICompatBaseUrl()
    const baseURL = managedModelBaseURL ?? configuredCompatBaseURL
    if (baseURL) {
      const headers = await getFirstPartyRequestHeaders(
        args.apiKey
          ? {
              apiKey: args.apiKey,
              includeApiKeyHeader: true,
            }
          : {},
      )
      const fetch = getWrappedClientFetch(args.fetchOverride, args.source)
      return new OpenAICompatInferenceClient({
        baseURL,
        headers,
        ...(fetch ? { fetch } : {}),
      })
    }
  }

  return new AnthropicInferenceClient(await getAnthropicClient(args))
}