config: centralize NanoBanana and audio defaults in defaults.go

Add internal/config/defaults.go with shared model IDs, output format, OpenAI/Gemini
audio literals, and speed defaults. Wire audio.DefaultProviderConfig, CLI flags,
GUI defaulting paths, and image package re-exports to these constants. Update
command_test to use the public config identifiers.

Made-with: Cursor

6 files changed, 63 insertions, 26 deletions

diff --git a/internal/audio/provider.go b/internal/audio/provider.go
index 0a9a4b8..424b946 100644
--- a/internal/audio/provider.go
+++ b/internal/audio/provider.go
@@ -5,6 +5,7 @@ import (
 	"fmt"
 	"strings"
 
+	"codeberg.org/snonux/totalrecall/internal/config"
 	"codeberg.org/snonux/totalrecall/internal/registry"
 )
 
@@ -117,19 +118,19 @@ func geminiAudioConfigFrom(c *Config) GeminiAudioConfig {
 	}
 }
 
-// DefaultConfig returns default configuration
+// DefaultProviderConfig returns default configuration (shared literals live in
+// internal/config/defaults.go).
 func DefaultProviderConfig() *Config {
 	return &Config{
-		Provider:     "gemini",
-		OutputDir:    "./",
-		OutputFormat: "mp3",
-		OpenAIModel:  "gpt-4o-mini-tts", // New model with voice instructions support
-		OpenAIVoice:  "alloy",
-		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-preview-tts",
-		GeminiSpeed:       1.0,
+		Provider:          config.ProviderGemini,
+		OutputDir:         "./",
+		OutputFormat:      config.DefaultAudioOutputFormat,
+		OpenAIModel:       config.DefaultOpenAIAudioModel,
+		OpenAIVoice:       config.DefaultOpenAIVoice,
+		OpenAISpeed:       config.DefaultOpenAIAudioSpeed,
+		OpenAIInstruction: config.DefaultOpenAIAudioInstruction,
+		GeminiTTSModel:    config.DefaultGeminiTTSModel,
+		GeminiSpeed:       config.DefaultGeminiAudioSpeed,
 	}
 }
 
diff --git a/internal/cli/command_test.go b/internal/cli/command_test.go
index afc1e18..80626eb 100644
--- a/internal/cli/command_test.go
+++ b/internal/cli/command_test.go
@@ -11,6 +11,7 @@ import (
 	"github.com/spf13/viper"
 
 	"codeberg.org/snonux/totalrecall/internal/audio"
+	"codeberg.org/snonux/totalrecall/internal/config"
 )
 
 func TestCreateRootCommand(t *testing.T) {
@@ -506,10 +507,10 @@ func TestMarkExplicitFlagValues(t *testing.T) {
 	if err := cmd.Flags().Set("format", "mp3"); err != nil {
 		t.Fatalf("Failed to set format flag: %v", err)
 	}
-	if err := cmd.Flags().Set("nanobanana-model", defaultNanoBananaModel); err != nil {
+	if err := cmd.Flags().Set("nanobanana-model", config.DefaultNanoBananaModel); err != nil {
 		t.Fatalf("Failed to set nanobanana-model flag: %v", err)
 	}
-	if err := cmd.Flags().Set("nanobanana-text-model", defaultNanoBananaTextModel); err != nil {
+	if err := cmd.Flags().Set("nanobanana-text-model", config.DefaultNanoBananaTextModel); err != nil {
 		t.Fatalf("Failed to set nanobanana-text-model flag: %v", err)
 	}
 
diff --git a/internal/cli/flags.go b/internal/cli/flags.go
index a343767..21e89d7 100644
--- a/internal/cli/flags.go
+++ b/internal/cli/flags.go
@@ -1,10 +1,8 @@
 package cli
 
-import "codeberg.org/snonux/totalrecall/internal/audio"
-
-const (
-	defaultNanoBananaModel     = "gemini-3.1-flash-image-preview"
-	defaultNanoBananaTextModel = "gemini-2.5-flash"
+import (
+	"codeberg.org/snonux/totalrecall/internal/audio"
+	"codeberg.org/snonux/totalrecall/internal/config"
 )
 
 // Flags holds all command-line flag values
@@ -84,7 +82,7 @@ func NewFlags() *Flags {
 		OpenAIImageQuality:  "standard",
 		OpenAIImageStyle:    "natural",
 		GeminiTTSModel:      defaults.GeminiTTSModel,
-		NanoBananaModel:     defaultNanoBananaModel,
-		NanoBananaTextModel: defaultNanoBananaTextModel,
+		NanoBananaModel:     config.DefaultNanoBananaModel,
+		NanoBananaTextModel: config.DefaultNanoBananaTextModel,
 	}
 }
diff --git a/internal/config/defaults.go b/internal/config/defaults.go
new file mode 100644
index 0000000..449e62a
--- /dev/null
+++ b/internal/config/defaults.go
@@ -0,0 +1,36 @@
+package config
+
+// Central defaults for Nano Banana (Gemini image) models and shared audio settings.
+// CLI, GUI, internal/audio, and callers resolve flag/config zero values against these
+// constants so model IDs and format strings are not duplicated (DRY).
+
+const (
+	// DefaultNanoBananaModel is the Gemini image model used for Nano Banana generation.
+	DefaultNanoBananaModel = "gemini-3.1-flash-image-preview"
+
+	// DefaultNanoBananaTextModel is the Gemini text model used for Nano Banana prompt
+	// and scene-related text generation.
+	DefaultNanoBananaTextModel = "gemini-2.5-flash"
+
+	// DefaultAudioOutputFormat is the default TTS file extension / container (e.g. mp3, wav).
+	DefaultAudioOutputFormat = "mp3"
+
+	// DefaultOpenAIAudioModel is the default OpenAI TTS model when that provider is selected.
+	DefaultOpenAIAudioModel = "gpt-4o-mini-tts"
+
+	// DefaultOpenAIVoice is the default OpenAI TTS voice.
+	DefaultOpenAIVoice = "alloy"
+
+	// DefaultGeminiTTSModel is the default Gemini TTS model identifier.
+	DefaultGeminiTTSModel = "gemini-2.5-flash-preview-tts"
+
+	// DefaultOpenAIAudioInstruction is the default system instruction for OpenAI TTS
+	// when generating Bulgarian learner audio.
+	DefaultOpenAIAudioInstruction = "You are speaking Bulgarian language (български език). Pronounce the Bulgarian text with authentic Bulgarian phonetics, not Russian. Speak slowly and clearly for language learners."
+)
+
+// Default audio tuning (non-string defaults for internal/audio.Config).
+const (
+	DefaultOpenAIAudioSpeed = 1.0
+	DefaultGeminiAudioSpeed = 1.0
+)
diff --git a/internal/gui/app.go b/internal/gui/app.go
index 56942e3..ff23acf 100644
--- a/internal/gui/app.go
+++ b/internal/gui/app.go
@@ -175,8 +175,8 @@ func DefaultConfig() *Config {
 		OutputDir:           outputDir,
 		AudioFormat:         audioDefaults.OutputFormat,
 		AudioProvider:       audioDefaults.Provider,
-		NanoBananaModel:     image.DefaultNanoBananaModel,
-		NanoBananaTextModel: image.DefaultNanoBananaTextModel,
+		NanoBananaModel:     appconfig.DefaultNanoBananaModel,
+		NanoBananaTextModel: appconfig.DefaultNanoBananaTextModel,
 		GeminiTTSModel:      audioDefaults.GeminiTTSModel,
 		ImageProvider:       image.ImageProviderNanoBanana,
 		TranslationProvider: translation.ProviderGemini,
@@ -249,7 +249,7 @@ func applyConfigDefaults(config *Config) *Config {
 		if strings.EqualFold(config.AudioProvider, "gemini") {
 			config.AudioFormat = defaults.AudioFormat
 		} else {
-			config.AudioFormat = "mp3"
+			config.AudioFormat = appconfig.DefaultAudioOutputFormat
 		}
 	}
 	if config.ImageProvider == "" {
@@ -344,7 +344,7 @@ func audioConfigForApp(config *Config) *audio.Config {
 		if provider == "gemini" {
 			outputFormat = defaults.OutputFormat
 		} else {
-			outputFormat = "mp3"
+			outputFormat = appconfig.DefaultAudioOutputFormat
 		}
 	}
 
diff --git a/internal/image/nanobanana.go b/internal/image/nanobanana.go
index 34a83ba..7694069 100644
--- a/internal/image/nanobanana.go
+++ b/internal/image/nanobanana.go
@@ -17,15 +17,16 @@ import (
 
 	"google.golang.org/genai"
 
+	"codeberg.org/snonux/totalrecall/internal/config"
 	"codeberg.org/snonux/totalrecall/internal/httpctx"
 )
 
 const (
 	// DefaultNanoBananaModel is the Gemini image model used for Nano Banana generation.
-	DefaultNanoBananaModel = "gemini-3.1-flash-image-preview"
+	DefaultNanoBananaModel = config.DefaultNanoBananaModel
 
 	// DefaultNanoBananaTextModel is the Gemini text model used for translation and scene generation.
-	DefaultNanoBananaTextModel = "gemini-2.5-flash"
+	DefaultNanoBananaTextModel = config.DefaultNanoBananaTextModel
 
 	nanoBananaAspectRatio = "4:3"
 	nanoBananaDataPrefix  = "data:image/png;base64,"