diff --git a/server/reflector/processors/__init__.py b/server/reflector/processors/__init__.py
index da890513..8a926f30 100644
--- a/server/reflector/processors/__init__.py
+++ b/server/reflector/processors/__init__.py
@@ -1,5 +1,6 @@
 from .base import Processor, ThreadedProcessor, Pipeline  # noqa: F401
 from .types import AudioFile, Transcript, Word, TitleSummary, FinalSummary  # noqa: F401
+from .audio_file_writer import AudioFileWriterProcessor  # noqa: F401
 from .audio_chunker import AudioChunkerProcessor  # noqa: F401
 from .audio_merge import AudioMergeProcessor  # noqa: F401
 from .audio_transcript import AudioTranscriptProcessor  # noqa: F401
diff --git a/server/reflector/processors/audio_file_writer.py b/server/reflector/processors/audio_file_writer.py
new file mode 100644
index 00000000..c597f81d
--- /dev/null
+++ b/server/reflector/processors/audio_file_writer.py
@@ -0,0 +1,35 @@
+from reflector.processors.base import Processor
+import av
+import wave
+from pathlib import Path
+
+
+class AudioFileWriterProcessor(Processor):
+    """
+    Write audio frames to a file.
+    """
+
+    INPUT_TYPE = av.AudioFrame
+    OUTPUT_TYPE = av.AudioFrame
+
+    def __init__(self, path: Path | str):
+        super().__init__()
+        if isinstance(path, str):
+            path = Path(path)
+        self.path = path
+        self.fd = None
+
+    async def _push(self, data: av.AudioFrame):
+        if not self.fd:
+            self.path.parent.mkdir(parents=True, exist_ok=True)
+            self.fd = wave.open(self.path.as_posix(), "wb")
+            self.fd.setnchannels(len(data.layout.channels))
+            self.fd.setsampwidth(data.format.bytes)
+            self.fd.setframerate(data.sample_rate)
+        self.fd.writeframes(data.to_ndarray().tobytes())
+        await self.emit(data)
+
+    async def _flush(self):
+        if self.fd:
+            self.fd.close()
+            self.fd = None
diff --git a/server/reflector/settings.py b/server/reflector/settings.py
index e776875b..0787b466 100644
--- a/server/reflector/settings.py
+++ b/server/reflector/settings.py
@@ -9,6 +9,9 @@ class Settings(BaseSettings):
     # Database
     DATABASE_URL: str = "sqlite:///./reflector.sqlite3"
 
+    # local data directory (audio for no)
+    DATA_DIR: str = "./data"
+
     # Whisper
     WHISPER_MODEL_SIZE: str = "tiny"
     WHISPER_REAL_TIME_MODEL_SIZE: str = "tiny"
diff --git a/server/reflector/views/rtc_offer.py b/server/reflector/views/rtc_offer.py
index aef00580..c0944a82 100644
--- a/server/reflector/views/rtc_offer.py
+++ b/server/reflector/views/rtc_offer.py
@@ -7,12 +7,14 @@ from reflector.logger import logger
 from aiortc import RTCPeerConnection, RTCSessionDescription, MediaStreamTrack
 from json import loads, dumps
 from enum import StrEnum
+from pathlib import Path
 import av
 from reflector.processors import (
     Pipeline,
     AudioChunkerProcessor,
     AudioMergeProcessor,
     AudioTranscriptAutoProcessor,
+    AudioFileWriterProcessor,
     TranscriptLinerProcessor,
     TranscriptTopicDetectorProcessor,
     TranscriptFinalSummaryProcessor,
@@ -64,7 +66,11 @@ class PipelineEvent(StrEnum):
 
 
 async def rtc_offer_base(
-    params: RtcOffer, request: Request, event_callback=None, event_callback_args=None
+    params: RtcOffer,
+    request: Request,
+    event_callback=None,
+    event_callback_args=None,
+    audio_filename: Path | None = None,
 ):
     # build an rtc session
     offer = RTCSessionDescription(sdp=params.sdp, type=params.type)
@@ -151,14 +157,18 @@ async def rtc_offer_base(
 
     # create a context for the whole rtc transaction
     # add a customised logger to the context
-    ctx.pipeline = Pipeline(
+    processors = []
+    if audio_filename is not None:
+        processors += [AudioFileWriterProcessor(path=audio_filename)]
+    processors += [
         AudioChunkerProcessor(),
         AudioMergeProcessor(),
         AudioTranscriptAutoProcessor.as_threaded(callback=on_transcript),
         TranscriptLinerProcessor(),
         TranscriptTopicDetectorProcessor.as_threaded(callback=on_topic),
         TranscriptFinalSummaryProcessor.as_threaded(callback=on_final_summary),
-    )
+    ]
+    ctx.pipeline = Pipeline(*processors)
     # FIXME: warmup is not working well yet
     # await ctx.pipeline.warmup()
 
diff --git a/server/reflector/views/transcripts.py b/server/reflector/views/transcripts.py
index f2a8425e..6f952938 100644
--- a/server/reflector/views/transcripts.py
+++ b/server/reflector/views/transcripts.py
@@ -5,14 +5,20 @@ from fastapi import (
     WebSocket,
     WebSocketDisconnect,
 )
+from fastapi.responses import FileResponse
+from starlette.concurrency import run_in_threadpool
 from pydantic import BaseModel, Field
 from uuid import uuid4
 from datetime import datetime
 from fastapi_pagination import Page, paginate
 from reflector.logger import logger
 from reflector.db import database, transcripts
+from reflector.settings import settings
 from .rtc_offer import rtc_offer_base, RtcOffer, PipelineEvent
 from typing import Optional
+from pathlib import Path
+from tempfile import NamedTemporaryFile
+import av
 
 
 router = APIRouter()
@@ -81,6 +87,44 @@ class Transcript(BaseModel):
     def topics_dump(self, mode="json"):
         return [topic.model_dump(mode=mode) for topic in self.topics]
 
+    def convert_audio_to_mp3(self):
+        fn = self.audio_mp3_filename
+        if fn.exists():
+            return
+
+        logger.info(f"Converting audio to mp3: {self.audio_filename}")
+        inp = av.open(self.audio_filename.as_posix(), "r")
+
+        # create temporary file for mp3
+        with NamedTemporaryFile(suffix=".mp3", delete=False) as tmp:
+            out = av.open(tmp.name, "w")
+            stream = out.add_stream("mp3")
+            for frame in inp.decode(audio=0):
+                frame.pts = None
+                for packet in stream.encode(frame):
+                    out.mux(packet)
+            for packet in stream.encode(None):
+                out.mux(packet)
+            out.close()
+
+            # move temporary file to final location
+            Path(tmp.name).rename(fn)
+
+    def unlink(self):
+        self.data_path.unlink(missing_ok=True)
+
+    @property
+    def data_path(self):
+        return Path(settings.DATA_DIR) / self.id
+
+    @property
+    def audio_filename(self):
+        return self.data_path / "audio.wav"
+
+    @property
+    def audio_mp3_filename(self):
+        return self.data_path / "audio.mp3"
+
 
 class TranscriptController:
     async def get_all(self) -> list[Transcript]:
@@ -112,6 +156,10 @@ class TranscriptController:
             setattr(transcript, key, value)
 
     async def remove_by_id(self, transcript_id: str) -> None:
+        transcript = await self.get_by_id(transcript_id)
+        if not transcript:
+            return
+        transcript.unlink()
         query = transcripts.delete().where(transcripts.c.id == transcript_id)
         await database.execute(query)
 
@@ -202,8 +250,24 @@ async def transcript_get_audio(transcript_id: str):
     if not transcript:
         raise HTTPException(status_code=404, detail="Transcript not found")
 
-    # TODO: Implement audio generation
-    return HTTPException(status_code=500, detail="Not implemented")
+    if not transcript.audio_filename.exists():
+        raise HTTPException(status_code=404, detail="Audio not found")
+
+    return FileResponse(transcript.audio_filename, media_type="audio/wav")
+
+
+@router.get("/transcripts/{transcript_id}/audio/mp3")
+async def transcript_get_audio_mp3(transcript_id: str):
+    transcript = await transcripts_controller.get_by_id(transcript_id)
+    if not transcript:
+        raise HTTPException(status_code=404, detail="Transcript not found")
+
+    if not transcript.audio_filename.exists():
+        raise HTTPException(status_code=404, detail="Audio not found")
+
+    await run_in_threadpool(transcript.convert_audio_to_mp3)
+
+    return FileResponse(transcript.audio_mp3_filename, media_type="audio/mp3")
 
 
 @router.get("/transcripts/{transcript_id}/topics", response_model=list[TranscriptTopic])
@@ -371,4 +435,5 @@ async def transcript_record_webrtc(
         request,
         event_callback=handle_rtc_event,
         event_callback_args=transcript_id,
+        audio_filename=transcript.audio_filename,
     )
diff --git a/server/tests/test_transcripts_rtc_ws.py b/server/tests/test_transcripts_rtc_ws.py
index 70ee209b..f38728c2 100644
--- a/server/tests/test_transcripts_rtc_ws.py
+++ b/server/tests/test_transcripts_rtc_ws.py
@@ -70,11 +70,15 @@ async def dummy_llm():
 
 
 @pytest.mark.asyncio
-async def test_transcript_rtc_and_websocket(dummy_transcript, dummy_llm):
+async def test_transcript_rtc_and_websocket(tmpdir, dummy_transcript, dummy_llm):
     # goal: start the server, exchange RTC, receive websocket events
     # because of that, we need to start the server in a thread
     # to be able to connect with aiortc
 
+    from reflector.settings import settings
+
+    settings.DATA_DIR = Path(tmpdir)
+
     # start server
     host = "127.0.0.1"
     port = 1255
@@ -188,3 +192,13 @@ async def test_transcript_rtc_and_websocket(dummy_transcript, dummy_llm):
     resp = await ac.get(f"/transcripts/{tid}")
     assert resp.status_code == 200
     assert resp.json()["status"] == "ended"
+
+    # check that audio is available
+    resp = await ac.get(f"/transcripts/{tid}/audio")
+    assert resp.status_code == 200
+    assert resp.headers["Content-Type"] == "audio/wav"
+
+    # check that audio/mp3 is available
+    resp = await ac.get(f"/transcripts/{tid}/audio/mp3")
+    assert resp.status_code == 200
+    assert resp.headers["Content-Type"] == "audio/mp3"