diff --git a/core/backend/model_identity_modalities_test.go b/core/backend/model_identity_modalities_test.go
index cc77a87b8..39d7930e2 100644
--- a/core/backend/model_identity_modalities_test.go
+++ b/core/backend/model_identity_modalities_test.go
@@ -49,6 +49,7 @@ type recordingBackend struct {
 	// `ModelIdentity`; the spec below asserts they are populated independently.
 	soundGenModelField string
 	soundGenParams     map[string]string
+	soundGenOutput     []byte
 }
 
 func newRecordingBackend() *recordingBackend {
@@ -73,6 +74,11 @@ func (r *recordingBackend) SoundGeneration(_ context.Context, in *pb.SoundGenera
 	r.record("SoundGeneration", in)
 	r.soundGenModelField = in.GetModel()
 	r.soundGenParams = in.GetParams()
+	if r.soundGenOutput != nil {
+		if err := os.WriteFile(in.GetDst(), r.soundGenOutput, 0o600); err != nil {
+			return nil, err
+		}
+	}
 	return &pb.Result{Success: true}, nil
 }
 
diff --git a/core/backend/soundgeneration.go b/core/backend/soundgeneration.go
index 98fb64201..31ea0e7f2 100644
--- a/core/backend/soundgeneration.go
+++ b/core/backend/soundgeneration.go
@@ -7,6 +7,7 @@ import (
 	"path/filepath"
 	"time"
 
+	"github.com/google/uuid"
 	"github.com/mudler/LocalAI/core/config"
 	"github.com/mudler/LocalAI/core/trace"
 	"github.com/mudler/LocalAI/pkg/grpc/proto"
@@ -56,7 +57,9 @@ func SoundGeneration(
 		return "", nil, fmt.Errorf("failed creating audio directory: %s", err)
 	}
 
-	fileName := utils.GenerateUniqueFileName(audioDir, "sound_generation", ".wav")
+	// Format normalization can rename .wav to .flac, so existence of the
+	// .wav path cannot determine whether an earlier recording owns this stem.
+	fileName := utils.GenerateUniqueFileName(audioDir, "sound_generation_"+uuid.NewString(), ".wav")
 	filePath := filepath.Join(audioDir, fileName)
 	if filePath, err = filepath.Abs(filePath); err != nil {
 		return "", nil, fmt.Errorf("failed resolving sound generation path: %w", err)
diff --git a/core/backend/soundgeneration_history_test.go b/core/backend/soundgeneration_history_test.go
new file mode 100644
index 000000000..3888de03e
--- /dev/null
+++ b/core/backend/soundgeneration_history_test.go
@@ -0,0 +1,50 @@
+package backend_test
+
+import (
+	"context"
+	"os"
+	"path/filepath"
+
+	"github.com/mudler/LocalAI/core/backend"
+	"github.com/mudler/LocalAI/core/config"
+	"github.com/mudler/LocalAI/pkg/audio"
+	"github.com/mudler/LocalAI/pkg/system"
+
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+)
+
+var _ = Describe("sound generation history", func() {
+	It("preserves previous recordings after FLAC extension normalization", func() {
+		rec := newRecordingBackend()
+		loader := newRecordingLoader(rec)
+		appCfg := config.NewApplicationConfig(config.WithSystemState(&system.SystemState{}))
+		appCfg.GeneratedContentDir = GinkgoT().TempDir()
+		cfg := identityModelCfg()
+
+		generate := func(content []byte) string {
+			rec.soundGenOutput = content
+			path, _, err := backend.SoundGeneration(context.Background(), "music", nil, nil, nil, nil, nil, nil, "", "", nil, "", "", "", nil, nil, loader, appCfg, cfg)
+			Expect(err).ToNot(HaveOccurred())
+			// The endpoint detects the backend's format and renames its .wav
+			// destination. Only the FLAC signature is needed for detection.
+			finalPath, contentType := audio.NormalizeAudioFile(path)
+			Expect(contentType).To(Equal("audio/flac"))
+			Expect(filepath.Ext(finalPath)).To(Equal(".flac"))
+			return finalPath
+		}
+
+		firstAudio := []byte("fLaC first recording fixture")
+		secondAudio := []byte("fLaC second recording fixture")
+		firstPath := generate(firstAudio)
+		secondPath := generate(secondAudio)
+
+		Expect(secondPath).ToNot(Equal(firstPath))
+		first, err := os.ReadFile(firstPath)
+		Expect(err).ToNot(HaveOccurred())
+		Expect(first).To(Equal(firstAudio))
+		second, err := os.ReadFile(secondPath)
+		Expect(err).ToNot(HaveOccurred())
+		Expect(second).To(Equal(secondAudio))
+	})
+})
diff --git a/docs/content/features/text-to-audio.md b/docs/content/features/text-to-audio.md
index 98e2a3913..c8109ba57 100644
--- a/docs/content/features/text-to-audio.md
+++ b/docs/content/features/text-to-audio.md
@@ -384,6 +384,10 @@ curl http://localhost:8080/v1/sound-generation -H "Content-Type: application/jso
 
 The `/v1/sound-generation` endpoint is compatible with the [ElevenLabs sound generation API](https://elevenlabs.io/docs/api-reference/sound-generation) and can produce music, sound effects, and other audio content. It responds with a binary audio file and the appropriate `Content-Type` header (for example `audio/wav`, `audio/mpeg`, `audio/flac`, or `audio/ogg`). The request body is JSON and supports two usage modes.
 
+Each generated sound has a unique filename. The filename remains unique when
+LocalAI changes its extension to match the audio format, so later generations
+do not overwrite earlier recordings in the WebUI history.
+
 **Simple mode:**
 
 | Parameter        | Type     | Required | Description                                  |