Initializing... drag & drop files here
Supports: 3GPP
3GPP (.3gpp, interchangeable with .3gp) is the 3GPP mobile-phone video container — low-resolution clips that pair H.263, MPEG-4 Part 2, or H.264 video with AMR or AAC audio. AMR (.amr) is the Adaptive Multi-Rate speech codec built for mobile voice. This converter is an audio extraction: it pulls the audio track out of the 3GPP file and writes it as AMR, discarding the video. It is the right target only when the source holds a short voice recording — a voice memo, an MMS clip, or audio from an old phone.
| Property | Value |
|---|---|
| Container | 3GPP (ISO base media / MP4 family), .3gpp = .3gp |
| Standard | 3GPP TS 26.244 |
| Video codecs | H.263, MPEG-4 Part 2, H.264/AVC |
| Audio codecs | AMR-NB, AMR-WB, AAC |
| Typical use | Mobile-phone video capture, MMS clips |
| Best for | Small video clips from older 3G/UMTS handsets |
| Property | Value |
|---|---|
| Codec | Adaptive Multi-Rate (AMR-NB by default) |
| Standard | IETF RFC 4867 (storage), 3GPP TS 26.071 (codec) |
| Sample rate | AMR-NB 8 kHz; AMR-WB 16 kHz |
| Bitrate | AMR-NB 4.75–12.2 kbps (8 modes); AMR-WB 6.6–23.85 kbps (9 modes) |
| Channels | Mono only (single-channel #!AMR\n header) |
| Frequency range | AMR-NB ~200–3400 Hz (telephone band); AMR-WB ~50 Hz–7 kHz |
| Best for | Speech: voice memos, MMS, GSM/UMTS voice |
| Poor for | Music or general audio — high frequencies are lost |
.3gpp or .3gp file, or click "+ Add Files." You can queue several clips and convert them in one batch..amr file. Files are uploaded over an encrypted connection, processed on our servers, and deleted automatically after a few hours. No sign-up, no watermark.No. AMR is an audio-only speech format, so the converter extracts the audio track and discards the video entirely. If you need the picture, convert to a video format instead — AMR cannot carry frames.
It depends on the source audio. AMR-NB is narrowband mono at an 8 kHz sample rate and 4.75–12.2 kbps, so anything above the telephone band (~3.4 kHz) is discarded. Speech survives well; music and general audio come out muffled and "telephone-quality." If the 3GPP already stores AMR (common for phone recordings), the result is close to a re-wrap; otherwise it is a lossy re-encode that cannot add back fidelity.
Use AMR-NB (8 kHz) for maximum compatibility with old phones and MMS — it is what "AMR" usually means. Use AMR-WB (16 kHz, up to 23.85 kbps) only if your target player or device supports wideband AMR and you want clearer voice. The defaults on this page (MONO, 8000 Hz) produce AMR-NB.
AMR opens in VLC, QuickTime, and RealPlayer, and audio editors such as Audacity can import it. Many older phones play AMR natively because it is the GSM/UMTS voice codec. On modern desktops VLC is the most reliable choice.
No. AMR is mono, narrowband, and speech-optimized, so music is reproduced poorly. For a song or any general-audio track, convert to 3GPP to MP3 or 3GPP to WAV instead — both keep the full frequency range and stereo.
Size and native phone support. AMR-NB was purpose-built for speech at 4.75–12.2 kbps, so a spoken clip lands in a fraction of the space an MP3 of the same length would need, and many phones and MMS gateways play AMR without any extra codec. For music or anything that needs full frequency range and stereo, MP3 is better — but for a plain voice memo bound for a phone, AMR is smaller and more compatible.
No. AMR is a single-channel (mono) speech codec — its file header (#!AMR) defines one channel only, so a stereo source is downmixed to mono during extraction. That is fine for voice, which is effectively mono anyway. To preserve left/right separation, extract to 3GPP to WAV or 3GPP to MP3 instead.
Essentially yes. AMR-NB is the adaptive narrowband speech codec 3GPP standardized for GSM and UMTS voice calls, and AMR-WB (marketed as HD Voice) is its wideband successor. Extracting a 3GPP clip to AMR stores the audio with the very codec family designed for cellular speech — which is why it sounds natural for talking and poor for music.
Files are uploaded over an encrypted connection, processed on our servers, and deleted automatically after a few hours. There is no sign-up, no watermark, and files are never shared or made public. In our testing, a 30-second 3GPP voice clip extracted to a roughly 30 KB AMR-NB file at the default 8 kHz mono settings.