Free online GSM to WAV converter

A GSM recording may need conversion when an audio editor, media player, transcription program, or hardware device accepts WAV but not GSM. WAV is also easier to edit and archive, but conversion cannot restore speech detail removed by GSM compression.

What the GSM format is

GSM usually means GSM 06.10, a lossy speech codec developed for low-bitrate telephone audio. A .gsm file commonly contains mono speech sampled at 8,000 Hz, producing much smaller files than uncompressed PCM audio.

GSM recordings occur in Asterisk and other telephony systems, voicemail platforms, call-recording applications, and older dictation or communications software. Many .gsm files are raw codec streams with no header, so conversion software may need the sample rate and channel count supplied manually. Some files instead use a WAV or another container with GSM-encoded audio.

What the WAV format is

WAV is an audio container that commonly stores uncompressed PCM. For this conversion, the usual output is 16-bit PCM WAV, which is widely supported by audio editors, speech-recognition software, video tools, operating systems, and hardware players.

PCM WAV preserves the decoded GSM signal without applying another lossy compression stage. It requires substantially more storage: 8 kHz, mono, 16-bit PCM uses about 16 KB per second, compared with roughly 1.6 KB per second for a 13 kbit/s GSM stream. WAV does not improve audio already degraded by GSM encoding.

How to convert GSM to WAV

For a single file without installing software, use the GSM-to-WAV converter on 101convert.com. Select the conversion, upload the GSM file, start the process, and download the resulting WAV. Use a local tool instead for confidential call recordings, unsupported raw files, or batch conversion.

FFmpeg handles both raw GSM streams and files with recognized containers. For a raw GSM file with the usual telephony properties, run:

ffmpeg -f gsm -ar 8000 -ac 1 -i input.gsm -c:a pcm_s16le output.wav

This specifies raw GSM input at 8 kHz and mono, then writes 16-bit little-endian PCM. Replace the filenames as needed and quote names containing spaces. If the file has a container header, let FFmpeg detect it instead:

ffmpeg -i input.gsm -c:a pcm_s16le output.wav

For batch processing, run the command once for each file or place it in a shell script. To create a WAV at another sample rate, add an output option such as -ar 16000; this resamples the decoded signal and does not restore missing high-frequency detail.

SoX can also decode raw GSM when its properties are explicit:

sox -t gsm -r 8000 -c 1 input.gsm -b 16 output.wav

Install a SoX build with GSM support. The command writes a WAV containing 16-bit PCM; omit -t gsm only when SoX can identify the input format from a valid container or header.

Quality and compatibility limits

GSM is lossy, so the WAV contains a decoded version of the GSM audio, not the original uncompressed recording. Resampling an 8 kHz source to 44.1 or 48 kHz changes the playback format but cannot recreate frequencies removed during encoding.

If the result is silent, distorted, or plays too fast, the source may be a raw GSM stream being interpreted with incorrect properties. Retry with -ar 8000 -ac 1 in FFmpeg or -r 8000 -c 1 in SoX. Conversely, forcing -f gsm on a containerized file can cause incorrect decoding; inspect uncertain files first with:

ffprobe input.gsm

Some applications require a particular WAV format, such as 16 kHz mono PCM for speech recognition or 44.1/48 kHz PCM for video work. Convert to that rate only after decoding, and verify the output codec, sample rate, channels, and bit depth in the receiving application.