mirror of
https://github.com/hanzo-java/sdk.git
synced 2026-11-20 20:02:20 +00:00
Regenerated with the java row's packages moved up a level: `cloud` under ai.hanzo named nothing a caller needed. ApiClient, JSON and the rest of the invoker sit in ai.hanzo beside the hand-written Hanzo; the examples are ai.hanzo.examples. An import of ai.hanzo.cloud.* becomes ai.hanzo.*. The artifact keeps its coordinate, ai.hanzo:hanzo-java-cloud. Co-authored-by: Hanzo Dev <[email protected]>
63 lines
2.4 KiB
Java
63 lines
2.4 KiB
Java
package ai.hanzo.examples;
|
|
|
|
import ai.hanzo.Hanzo;
|
|
import ai.hanzo.ApiException;
|
|
import ai.hanzo.api.AiApi;
|
|
import ai.hanzo.model.OpenaiChatCompletionChoice;
|
|
import ai.hanzo.model.OpenaiChatCompletionMessage;
|
|
import ai.hanzo.model.OpenaiChatCompletionRequest;
|
|
import ai.hanzo.model.PostChat200Response;
|
|
|
|
import java.util.List;
|
|
|
|
/**
|
|
* chat — one completion.
|
|
*
|
|
* <p>Operation: {@code post_chat_completions} — POST /v1/chat/completions,
|
|
* the gateway's own inference route.
|
|
*
|
|
* <p>Non-streaming on purpose: streaming is SSE, a different transport that a
|
|
* generated client hands back as an opaque body, so demonstrating it here would
|
|
* teach the wrong shape.
|
|
*
|
|
* <p>The document types the route — {@code OpenaiChatCompletionRequest} in,
|
|
* {@code PostChat200Response} out (the completion plus the {@code routing} the
|
|
* gateway chose) — so the prompt goes in as a value
|
|
* and the reply is {@code choices[0].message.content}.
|
|
*
|
|
* <pre>
|
|
* HANZO_API_KEY=hk-... ./gradlew :examples:chat
|
|
* </pre>
|
|
*/
|
|
public final class Chat {
|
|
|
|
/** A model the gateway serves. {@code HANZO_MODEL} overrides it. */
|
|
private static final String DEFAULT_MODEL = "zen5";
|
|
|
|
public static void main(String[] args) {
|
|
AiApi ai = new AiApi(Hanzo.client());
|
|
String model = System.getenv("HANZO_MODEL");
|
|
if (model == null || model.trim().isEmpty()) {
|
|
model = DEFAULT_MODEL;
|
|
}
|
|
try {
|
|
// The two nulls are the optional X-Max-Cost and X-Max-Latency-Ms
|
|
// routing bounds; unset, the gateway routes without a ceiling.
|
|
PostChat200Response reply = ai.postChatCompletions(new OpenaiChatCompletionRequest()
|
|
.model(model)
|
|
.messages(List.of(new OpenaiChatCompletionMessage()
|
|
.role("user")
|
|
.content("Say hello in one short sentence."))), null, null);
|
|
List<OpenaiChatCompletionChoice> choices = reply.getChoices();
|
|
String content = choices == null || choices.isEmpty() || choices.get(0).getMessage() == null
|
|
? ""
|
|
: choices.get(0).getMessage().getContent();
|
|
System.out.printf("model %s%n", reply.getModel());
|
|
System.out.printf("reply %s%n", content);
|
|
} catch (ApiException e) {
|
|
System.err.printf("chat failed: HTTP %d %s%n", e.getCode(), e.getResponseBody());
|
|
System.exit(1);
|
|
}
|
|
}
|
|
}
|