Skip to content

The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.

[Partner Nodes] feat(FishAudio): add Drama 3 model (#16999)

Alexander Piskun8b1a8a3cd546
diff --git a/comfy_api_nodes/nodes_fishaudio.py b/comfy_api_nodes/nodes_fishaudio.py
index 3d30190e6..331c4c7ab 100644
--- a/comfy_api_nodes/nodes_fishaudio.py
+++ b/comfy_api_nodes/nodes_fishaudio.py
@@ -37,6 +37,8 @@ FISHAUDIO_VOICES = [
("5212eb29e500460391d03af42af6552e", "Warm Conversational Voice (en)"),
("5c8dc6a69c0b4edfb32634db6384bf34", "Warm Storyteller (en)"),
("7a18a1851d2649108c48ec9f2c80eb2c", "Dramatic Character Male (en)"),
+ ("c70cfb98867a4bfb9ae20d2401257b92", "Delia (en)"),
+ ("f3a2b90078d54a65af4b95f63bc7798e", "Sheila (en)"),
("59cb5986671546eaa6ca8ae6f29f6d22", "News Narrator (zh)"),
("bf6c479f5a384b8d857310030035824b", "Lively Female (zh)"),
("faccba1a8ac54016bcfc02761285e67f", "Gentle Female (zh)"),
@@ -123,7 +125,7 @@ def _multi_speaker_inputs() -> list:
min=0,
max=5,
),
- tooltip="Voices for synthesis. Leave empty for the default voice. "
+ tooltip="Voices for synthesis. Leave empty for a random voice. "
"With two or more voices, mark speaker changes in the text with @Voice1, @Voice2, etc.",
),
*_tts_option_inputs(),
@@ -186,7 +188,8 @@ class FishAudioTextToSpeech(IO.ComfyNode):
display_name="Fish Audio Text to Speech",
category="partner/audio/Fish Audio",
description="Convert text to speech. Supports emotion cues in the text "
- "([happy], [whispering] on s2.1-pro; (happy) on s1) and multi-speaker dialogue "
+ "([happy], [whispering] on s2.1-pro; (happy) on s1), plain-language [directions] and "
+ "`<whisper>...</whisper>` word tags on drama-3, and multi-speaker dialogue "
"via @Voice1/@Voice2 tags with multiple connected voices.",
inputs=[
IO.String.Input(
@@ -199,6 +202,7 @@ class FishAudioTextToSpeech(IO.ComfyNode):
IO.DynamicCombo.Input(
"model",
options=[
+ IO.DynamicCombo.Option("drama-3", _multi_speaker_inputs()),
IO.DynamicCombo.Option("s2.1-pro", _multi_speaker_inputs()),
IO.DynamicCombo.Option(
"s1",
@@ -206,7 +210,7 @@ class FishAudioTextToSpeech(IO.ComfyNode):
IO.Custom(FISHAUDIO_VOICE).Input(
"voice",
optional=True,
- tooltip="Voice for synthesis. Leave unconnected for the default voice.",
+ tooltip="Voice for synthesis. Leave unconnected for a random voice.",
),
*_tts_option_inputs(),
],
@@ -235,16 +239,19 @@ class FishAudioTextToSpeech(IO.ComfyNode):
],
is_api_node=True,
price_badge=IO.PriceBadge(
- depends_on=IO.PriceBadgeDepends(widgets=["text"]),
+ depends_on=IO.PriceBadgeDepends(widgets=["text", "model"]),
expr="""
(
$t := widgets.text;
+ $rate := widgets.model = "drama-3"
+ ? ($now() < "2026-11-13T00:00:00.000Z" ? 14.3 : 35.75)
+ : 21.45;
$type($t) = "string"
? (
$bytes := $length($t) + 2 * $count($match($t, /[^\\x00-\\x7F]/));
- {"type":"usd","usd": $bytes * 21.45 / 1000000, "format":{"approximate":true}}
+ {"type":"usd","usd": $bytes * $rate / 1000000, "format":{"approximate":true}}
)
- : {"type":"usd","usd": 0.02145, "format":{"approximate":true, "suffix":"/1K bytes"}}
+ : {"type":"usd","usd": $rate / 1000, "format":{"approximate":true, "suffix":"/1K bytes"}}
)
""",
),