diff --git a/Cargo.lock b/Cargo.lock
index a6bf75f..88facb8 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -2831,7 +2831,7 @@ dependencies = [
 
 [[package]]
 name = "wavekat-vad"
-version = "0.1.8"
+version = "0.1.9"
 dependencies = [
  "criterion",
  "hound",
diff --git a/README.md b/README.md
index 2ecab27..d571ae2 100644
--- a/README.md
+++ b/README.md
@@ -42,13 +42,14 @@ wavekat-vad = { version = "0.1", features = ["webrtc", "silero", "ten-vad", "fir
 Performance measured against the [TEN-VAD testset](https://github.com/TEN-framework/ten-vad/tree/main/testset) — 30 audio files from LibriSpeech, GigaSpeech, and DNS Challenge with manual speech/non-speech annotations. Threshold: 0.5.
 
 <!-- benchmark-table-start -->
-*v0.1.8*
+*v0.1.9*
 
 | Backend | Precision | Recall | F1 Score | Frame Size | Avg Inference | RTF |
 |---------|-----------|--------|----------|------------|---------------|-----|
-| WebRTC | 0.821 | 0.983 | 0.895 | 480 (30 ms) | 2.6 µs | 0.0001 |
-| Silero | 0.938 | 0.938 | 0.938 | 512 (32 ms) | 121.0 µs | 0.0038 |
-| TEN-VAD | 0.942 | 0.915 | 0.928 | 256 (16 ms) | 61.7 µs | 0.0039 |
+| WebRTC | 0.821 | 0.983 | 0.895 | 480 (30 ms) | 2.7 µs | 0.0001 |
+| Silero | 0.938 | 0.938 | 0.938 | 512 (32 ms) | 115.4 µs | 0.0036 |
+| TEN-VAD | 0.942 | 0.915 | 0.928 | 256 (16 ms) | 59.6 µs | 0.0037 |
+| FireRedVAD | 0.950 | 0.879 | 0.913 | 160 (10 ms) | 533.9 µs | 0.0534 |
 <!-- benchmark-table-end -->
 
 > Accuracy metrics are deterministic; inference times are approximate and vary by hardware. Measured with `--release` on GitHub Actions `ubuntu-latest` runners. Run locally: `make accuracy` or `make bench`
diff --git a/crates/wavekat-vad/CHANGELOG.md b/crates/wavekat-vad/CHANGELOG.md
index 0bc0ce6..bc4780f 100644
--- a/crates/wavekat-vad/CHANGELOG.md
+++ b/crates/wavekat-vad/CHANGELOG.md
@@ -7,6 +7,12 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.1.9](https://github.com/wavekat/wavekat-vad/compare/v0.1.8...v0.1.9) - 2026-03-25
+
+### Added
+
+- add FireRedVAD backend ([#38](https://github.com/wavekat/wavekat-vad/pull/38))
+
 ## [0.1.8](https://github.com/wavekat/wavekat-vad/compare/v0.1.7...v0.1.8) - 2026-03-24
 
 ### Other
diff --git a/crates/wavekat-vad/Cargo.toml b/crates/wavekat-vad/Cargo.toml
index 03639f0..8538686 100644
--- a/crates/wavekat-vad/Cargo.toml
+++ b/crates/wavekat-vad/Cargo.toml
@@ -1,6 +1,6 @@
 [package]
 name = "wavekat-vad"
-version = "0.1.8"
+version = "0.1.9"
 edition = "2021"
 description = "Unified voice activity detection with multiple backends"
 license = "Apache-2.0"