docs: add full vLLM-Omni capability examples

Expand the English and Korean deployment guides with capability mappings and examples for speaker-conditioned TTS, TextQA, SpeechQA, multimodal chat, and SpokenQA.

README.md

@@ -275,6 +275,16 @@ compatible `feat/axk2_raon_speech` branch. The validated revision is
 you can optionally pin it to match the vLLM-Omni implementation revision used
 for validation.
 
+The online server supports these request patterns:
+
+| Capability | Request | Response |
+| --- | --- | --- |
+| Direct or speaker-conditioned TTS / voice cloning | Text, optionally with reference audio | WAV from `/v1/audio/speech` |
+| STT | Audio with a written transcription instruction | Text from `/v1/chat/completions` |
+| TextQA | Written question | Text answer from `/v1/chat/completions` |
+| SpeechQA / multimodal chat | Audio context with a written question or instruction | Text answer from `/v1/chat/completions` |
+| SpokenQA / SpeechChat | Spoken audio question without a written prompt | Text answer from `/v1/chat/completions` |
+
 ### Install and serve
 
 ```bash
@@ -330,6 +340,26 @@ curl -fS -X POST http://localhost:8091/v1/audio/speech \
 EOF
 ```
 
+### Speaker-conditioned TTS / voice cloning
+
+Encode the reference speaker audio as a portable single-line base64 data URL:
+
+```bash
+REF_AUDIO_B64="$(base64 < speaker_ref.wav | tr -d '\n')"
+
+curl -fS -X POST http://localhost:8091/v1/audio/speech \
+    -H "Content-Type: application/json" \
+    --data-binary @- \
+    --output axk2_conditioned.wav <<EOF
+{
+  "model": "KRAFTON/A.X-K2-Raon-Speech-21B-A3B",
+  "input": "Hello, this sentence uses the reference speaker.",
+  "ref_audio": "data:audio/wav;base64,$REF_AUDIO_B64",
+  "response_format": "wav"
+}
+EOF
+```
+
 ### STT
 
 Encode the input audio as a portable single-line base64 data URL:
@@ -363,6 +393,93 @@ curl -fS -X POST http://localhost:8091/v1/chat/completions \
 EOF
 ```
 
+### TextQA
+
+Send a text-only question:
+
+```bash
+curl -fS -X POST http://localhost:8091/v1/chat/completions \
+    -H "Content-Type: application/json" \
+    --data-binary @- <<'EOF'
+{
+  "model": "KRAFTON/A.X-K2-Raon-Speech-21B-A3B",
+  "messages": [
+    {
+      "role": "user",
+      "content": [
+        {
+          "type": "text",
+          "text": "What is the capital of South Korea?"
+        }
+      ]
+    }
+  ]
+}
+EOF
+```
+
+### SpeechQA / multimodal chat
+
+Use the audio as context and include a written question:
+
+```bash
+AUDIO_B64="$(base64 < audio.wav | tr -d '\n')"
+
+curl -fS -X POST http://localhost:8091/v1/chat/completions \
+    -H "Content-Type: application/json" \
+    --data-binary @- <<EOF
+{
+  "model": "KRAFTON/A.X-K2-Raon-Speech-21B-A3B",
+  "messages": [
+    {
+      "role": "user",
+      "content": [
+        {
+          "type": "audio_url",
+          "audio_url": {
+            "url": "data:audio/wav;base64,$AUDIO_B64"
+          }
+        },
+        {
+          "type": "text",
+          "text": "What is the speaker talking about?"
+        }
+      ]
+    }
+  ]
+}
+EOF
+```
+
+### SpokenQA / SpeechChat
+
+Send a spoken question without a written prompt:
+
+```bash
+QUESTION_AUDIO_B64="$(base64 < question.wav | tr -d '\n')"
+
+curl -fS -X POST http://localhost:8091/v1/chat/completions \
+    -H "Content-Type: application/json" \
+    --data-binary @- <<EOF
+{
+  "model": "KRAFTON/A.X-K2-Raon-Speech-21B-A3B",
+  "messages": [
+    {
+      "role": "user",
+      "content": [
+        {
+          "type": "audio_url",
+          "audio_url": {
+            "url": "data:audio/wav;base64,$QUESTION_AUDIO_B64"
+          }
+        }
+      ]
+    }
+  ]
+}
+EOF
+```
+
 ## Training
 
 The released checkpoint is the final model following:

README_ko.md

@@ -275,6 +275,16 @@ KRAFTON의 공식
 [`7bcaf21`](https://github.com/krafton-ai/vllm-omni/commit/7bcaf21df881bda4280cbbf9752fc356e753d3bb)이며,
 검증에 사용된 vLLM-Omni 구현 리비전과 일치시키려면 이 커밋으로 고정하세요.
 
+온라인 서버는 다음 요청 형식을 지원합니다.
+
+| 기능 | 요청 | 응답 |
+| --- | --- | --- |
+| 기본 또는 화자 조건 TTS / 음성 복제 | 텍스트와 선택적 참조 오디오 | `/v1/audio/speech`에서 WAV |
+| STT | 오디오와 텍스트 전사 지시문 | `/v1/chat/completions`에서 텍스트 |
+| TextQA | 텍스트 질문 | `/v1/chat/completions`에서 텍스트 답변 |
+| SpeechQA / 멀티모달 대화 | 오디오 맥락과 텍스트 질문 또는 지시문 | `/v1/chat/completions`에서 텍스트 답변 |
+| SpokenQA / SpeechChat | 텍스트 프롬프트 없이 음성으로 전달한 질문 | `/v1/chat/completions`에서 텍스트 답변 |
+
 ### 설치 및 실행
 
 ```bash
@@ -329,6 +339,27 @@ curl -fS -X POST http://localhost:8091/v1/audio/speech \
 EOF
 ```
 
+### 화자 조건 TTS / 음성 복제
+
+참조 화자의 오디오를 이식성이 높은 한 줄짜리 base64 데이터 URL로
+인코딩합니다.
+
+```bash
+REF_AUDIO_B64="$(base64 < speaker_ref.wav | tr -d '\n')"
+
+curl -fS -X POST http://localhost:8091/v1/audio/speech \
+    -H "Content-Type: application/json" \
+    --data-binary @- \
+    --output axk2_conditioned.wav <<EOF
+{
+  "model": "KRAFTON/A.X-K2-Raon-Speech-21B-A3B",
+  "input": "Hello, this sentence uses the reference speaker.",
+  "ref_audio": "data:audio/wav;base64,$REF_AUDIO_B64",
+  "response_format": "wav"
+}
+EOF
+```
+
 ### STT
 
 입력 오디오를 이식성이 높은 한 줄짜리 base64 데이터 URL로 인코딩합니다.
@@ -362,6 +393,93 @@ curl -fS -X POST http://localhost:8091/v1/chat/completions \
 EOF
 ```
 
+### TextQA
+
+텍스트로만 구성된 질문을 전송합니다.
+
+```bash
+curl -fS -X POST http://localhost:8091/v1/chat/completions \
+    -H "Content-Type: application/json" \
+    --data-binary @- <<'EOF'
+{
+  "model": "KRAFTON/A.X-K2-Raon-Speech-21B-A3B",
+  "messages": [
+    {
+      "role": "user",
+      "content": [
+        {
+          "type": "text",
+          "text": "What is the capital of South Korea?"
+        }
+      ]
+    }
+  ]
+}
+EOF
+```
+
+### SpeechQA / 멀티모달 대화
+
+오디오를 맥락으로 사용하고 텍스트 질문을 함께 전송합니다.
+
+```bash
+AUDIO_B64="$(base64 < audio.wav | tr -d '\n')"
+
+curl -fS -X POST http://localhost:8091/v1/chat/completions \
+    -H "Content-Type: application/json" \
+    --data-binary @- <<EOF
+{
+  "model": "KRAFTON/A.X-K2-Raon-Speech-21B-A3B",
+  "messages": [
+    {
+      "role": "user",
+      "content": [
+        {
+          "type": "audio_url",
+          "audio_url": {
+            "url": "data:audio/wav;base64,$AUDIO_B64"
+          }
+        },
+        {
+          "type": "text",
+          "text": "What is the speaker talking about?"
+        }
+      ]
+    }
+  ]
+}
+EOF
+```
+
+### SpokenQA / SpeechChat
+
+텍스트 프롬프트 없이 음성 질문을 전송합니다.
+
+```bash
+QUESTION_AUDIO_B64="$(base64 < question.wav | tr -d '\n')"
+
+curl -fS -X POST http://localhost:8091/v1/chat/completions \
+    -H "Content-Type: application/json" \
+    --data-binary @- <<EOF
+{
+  "model": "KRAFTON/A.X-K2-Raon-Speech-21B-A3B",
+  "messages": [
+    {
+      "role": "user",
+      "content": [
+        {
+          "type": "audio_url",
+          "audio_url": {
+            "url": "data:audio/wav;base64,$QUESTION_AUDIO_B64"
+          }
+        }
+      ]
+    }
+  ]
+}
+EOF
+```
+
 ## Training
 
 공개된 checkpoint는 다음 과정을 거친 최종 모델입니다.