summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorPaul Buetow <paul@buetow.org>2026-04-01 13:51:09 +0300
committerPaul Buetow <paul@buetow.org>2026-04-01 13:51:09 +0300
commit9dbf30e9eacccfc23d980c94c3b4d91b5cdb3459 (patch)
tree87cc415bb4564d88a4355410c14cdf777e276758
parent155b3e78e184b8f1c2193d875d4db8a56ea36d34 (diff)
zf: align Gemini TTS default model
-rw-r--r--internal/audio/gemini_provider.go2
-rw-r--r--internal/audio/gemini_provider_test.go2
-rw-r--r--internal/audio/provider.go4
-rw-r--r--internal/audio/provider_test.go4
4 files changed, 8 insertions, 4 deletions
diff --git a/internal/audio/gemini_provider.go b/internal/audio/gemini_provider.go
index 8d0c517..9900005 100644
--- a/internal/audio/gemini_provider.go
+++ b/internal/audio/gemini_provider.go
@@ -14,7 +14,7 @@ import (
)
const (
- defaultGeminiTTSModel = "gemini-2.5-flash"
+ defaultGeminiTTSModel = "gemini-2.5-flash-preview-tts"
geminiTTSLanguageCode = "bg"
geminiTTSChannels = 1
geminiTTSSampleRate = 24000
diff --git a/internal/audio/gemini_provider_test.go b/internal/audio/gemini_provider_test.go
index b8ab4c9..1f2a25a 100644
--- a/internal/audio/gemini_provider_test.go
+++ b/internal/audio/gemini_provider_test.go
@@ -24,7 +24,7 @@ func TestNewGeminiProvider(t *testing.T) {
name: "valid config",
config: &Config{
GoogleAPIKey: "test-key",
- GeminiTTSModel: "gemini-2.5-flash",
+ GeminiTTSModel: "gemini-2.5-flash-preview-tts",
GeminiVoice: "Kore",
GeminiSpeed: 1.0,
},
diff --git a/internal/audio/provider.go b/internal/audio/provider.go
index 7ca4282..b961482 100644
--- a/internal/audio/provider.go
+++ b/internal/audio/provider.go
@@ -32,7 +32,7 @@ type Config struct {
// Gemini-specific settings
GoogleAPIKey string
- GeminiTTSModel string // "gemini-2.5-flash"
+ GeminiTTSModel string // "gemini-2.5-flash-preview-tts"
GeminiVoice string // Prebuilt Gemini TTS voice name, or empty for the model default
GeminiSpeed float64 // Prompt hint for desired speech speed
}
@@ -48,7 +48,7 @@ func DefaultProviderConfig() *Config {
OpenAISpeed: 1.0,
// OpenAISpeed: 0.98, // Default speed for clarity
OpenAIInstruction: "You are speaking Bulgarian language (български език). Pronounce the Bulgarian text with authentic Bulgarian phonetics, not Russian. Speak slowly and clearly for language learners.",
- GeminiTTSModel: "gemini-2.5-flash",
+ GeminiTTSModel: "gemini-2.5-flash-preview-tts",
GeminiSpeed: 1.0,
}
}
diff --git a/internal/audio/provider_test.go b/internal/audio/provider_test.go
index 24d2729..6b86e3c 100644
--- a/internal/audio/provider_test.go
+++ b/internal/audio/provider_test.go
@@ -50,6 +50,10 @@ func TestDefaultProviderConfig(t *testing.T) {
t.Errorf("Expected OpenAI speed 1.0, got %f", config.OpenAISpeed)
}
+ if config.GeminiTTSModel != "gemini-2.5-flash-preview-tts" {
+ t.Errorf("Expected Gemini TTS model 'gemini-2.5-flash-preview-tts', got '%s'", config.GeminiTTSModel)
+ }
+
}
func TestNewProvider(t *testing.T) {