Tetap teratur dengan koleksi
Simpan dan kategorikan konten berdasarkan preferensi Anda.
Respons sederhana berbentuk balon chat secara visual dan menggunakan text-to-speech
(TTS) atau Bahasa Markup Sintesis Ucapan (SSML) untuk suara. Dengan menggunakan
tanggapan sederhana dalam percakapan, Anda dapat membuat pengguna tetap terlibat dengan
dan antarmuka audio yang dapat dipasangkan
dengan elemen percakapan lainnya.
Konten balon chat dalam respons sederhana harus berupa subset fonetik atau bagian lengkap
transkrip output TTS/SSML. Hal ini membantu pengguna memetakan apa yang
dan meningkatkan
pemahaman dalam berbagai kondisi.
Dalam perintah, teks yang Anda berikan dalam objek first_simple dan last_simple
menggunakan properti respons sederhana. Asisten Google mengirimkan semua informasi
respons yang diberikan di prompt, lalu mengirimkan respons lengkap akhir dalam antrean prompt.
Properti
Jenis respons sederhana memiliki properti berikut:
Properti
Jenis
Persyaratan
Deskripsi
speech
string
Opsional
Mewakili kata yang akan diucapkan kepada pengguna dalam SSML atau text-to-speech.
Jika kolom override dalam perintah yang memuatnya adalah "true",
maka ucapan yang didefinisikan dalam isian ini menggantikan
ucapan prompt.
text
string
Opsional
Teks yang akan ditampilkan dalam balon chat. String yang berisi lebih dari 640 karakter
terpotong di jeda kata pertama (atau spasi kosong) sebelum 640
karakter. Sebaiknya gunakan kurang dari 300 karakter untuk mencegah
agar konten tidak memanjang melewati layar, terutama bila disambungkan dengan
{i>card <i}atau elemen visual lainnya.
Jika tidak disediakan, Asisten akan merender versi tampilan dari
Sebagai gantinya, kolom speech. Jika kolom override di
prompt yang memuatnya adalah "false", maka teks yang ditentukan dalam kolom ini adalah
yang ditambahkan ke teks prompt sederhana sebelumnya.
Kode contoh
YAML
candidates:-first_simple:variants:-speech:This is the first simple response.text:This is the 1st simple response.last_simple:variants:-speech:This is the last simple response.text:This is the last simple response.
JSON
{"candidates":[{"first_simple":{"variants":[{"speech":"This is the first simple response.","text":"This is the 1st simple response."}]},"last_simple":{"variants":[{"speech":"This is the last simple response.","text":"This is the last simple response."}]}}]}
Node.js
app.handle('Simple',conv=>{conv.add(newSimple({speech:'This is the first simple response.',text:'This is the 1st simple response.'}));conv.add(newSimple({speech:'This is the last simple response.',text:'This is the last simple response.'}));});
JSON
{"responseJson":{"session":{"id":"session_id","params":{}},"prompt":{"override":false,"firstSimple":{"speech":"This is the first simple response.","text":"This is the 1st simple response."},"lastSimple":{"speech":"This is the last simple response.","text":"This is the last simple response."}}}}
SSML dan suara
Gunakan SSML dan suara dalam respons Anda untuk membuatnya lebih rapi dan meningkatkan
{i>user experience<i}. Lihat dokumentasi SSML untuk informasi selengkapnya.
Koleksi audio
Kami menyediakan berbagai suara pendek dan gratis di koleksi suara kami. Ini
suara di-host untuk Anda, jadi yang perlu Anda lakukan hanyalah memasukkannya ke dalam SSML Anda.
[null,null,["Terakhir diperbarui pada 2025-07-26 UTC."],[[["\u003cp\u003eSimple responses use chat bubbles with text-to-speech or SSML for audio, keeping users engaged with a clear interface.\u003c/p\u003e\n"],["\u003cp\u003eChat bubble content should be a phonetic subset or a complete transcript of the audio output for better user comprehension.\u003c/p\u003e\n"],["\u003cp\u003eSimple responses are sent before the final rich response, and some surfaces may only display the rich response.\u003c/p\u003e\n"],["\u003cp\u003eThe \u003ccode\u003espeech\u003c/code\u003e property defines what the Assistant says, while the \u003ccode\u003etext\u003c/code\u003e property defines what's displayed in the chat bubble.\u003c/p\u003e\n"],["\u003cp\u003eSSML and sounds can be used to enhance simple responses, with a sound library available for use.\u003c/p\u003e\n"]]],[],null,["# Simple responses take the form of a chat bubble visually and use text-to-speech\n(TTS) or Speech Synthesis Markup Language (SSML) for sound. By using short\nsimple responses in conversation, you can keep users engaged with a clear visual\nand audio interface that can be paired with other conversational elements.\n\nChat bubble content in a simple response must be a phonetic subset or a complete\ntranscript of the TTS/SSML output. This helps users map out what your Action\nsays and increases comprehension in various conditions.\n\nIn a prompt, text you provide in the `first_simple` and `last_simple` objects\nuse the properties of a simple response. Google Assistant sends all simple\nresponses in a prompt, then sends the final rich response in the prompt queue.\n| **Note:** Some surfaces (like smart displays) only display one piece of content at a time instead of using the chat bubble format. As a result, only the rich response appears when you provide both a simple and a rich response.\n\nProperties\n----------\n\nThe simple response type has the following properties:\n\n| Property | Type | Requirement | Description |\n|----------|--------|-------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|\n| `speech` | string | Optional | Represents the words to be spoken to the user in SSML or text-to-speech. If the `override` field in the containing prompt is \"true\", then speech defined in this field replaces the previous simple prompt's speech. |\n| `text` | string | Optional | Text to display in the chat bubble. Strings longer than 640 characters are truncated at the first word break (or whitespace) before 640 characters. We recommend using less than 300 characters to prevent content from extending past the screen, especially when paired with a card or other visual element. If not provided, Assistant renders a display version of the `speech` field instead. If the `override` field in the containing prompt is \"false\", then text defined in this field is appended to the previous simple prompt's text. |\n\nSample code\n-----------\n\n### YAML\n\n```yaml\ncandidates:\n - first_simple:\n variants:\n - speech: This is the first simple response.\n text: This is the 1st simple response.\n last_simple:\n variants:\n - speech: This is the last simple response.\n text: This is the last simple response.\n```\n\n### JSON\n\n```gdscript\n{\n \"candidates\": [\n {\n \"first_simple\": {\n \"variants\": [\n {\n \"speech\": \"This is the first simple response.\",\n \"text\": \"This is the 1st simple response.\"\n }\n ]\n },\n \"last_simple\": {\n \"variants\": [\n {\n \"speech\": \"This is the last simple response.\",\n \"text\": \"This is the last simple response.\"\n }\n ]\n }\n }\n ]\n}\n```\n\n### Node.js\n\n```javascript\napp.handle('Simple', conv =\u003e {\n conv.add(new Simple({\n speech: 'This is the first simple response.',\n text: 'This is the 1st simple response.'\n }));\n conv.add(new Simple({\n speech: 'This is the last simple response.',\n text: 'This is the last simple response.'\n }));\n});\n```\n\n### JSON\n\n```carbon\n{\n \"responseJson\": {\n \"session\": {\n \"id\": \"session_id\",\n \"params\": {}\n },\n \"prompt\": {\n \"override\": false,\n \"firstSimple\": {\n \"speech\": \"This is the first simple response.\",\n \"text\": \"This is the 1st simple response.\"\n },\n \"lastSimple\": {\n \"speech\": \"This is the last simple response.\",\n \"text\": \"This is the last simple response.\"\n }\n }\n }\n}\n```\n\nSSML and sounds\n---------------\n\nUse SSML and sounds in your responses to give them more polish and enhance the\nuser experience. See the [SSML documentation](/assistant/conversational/ssml) for more information.\n\nSound library\n-------------\n\nWe provide a variety of free, short sounds in our [sound library](/assistant/tools/sound-library). These\nsounds are hosted for you, so all you need to do is include them in your SSML."]]