Enhanced vs Flow Matching: which AI audio restoration preset to use

The Refiner exposes two presets for AI audio restoration, and they are not just different settings on the same model — they are different restoration approaches. Enhanced uses neural reconstruction to synthesize high-frequency detail. Flow Matching uses compute-intensive, iterative reconstruction that historically took minutes and more credits. They produce different kinds of output and fail in different ways. Flow Matching is currently paused for new jobs while its output quality is reviewed.

This is the long-form decision guide. There is a shorter technical summary on the presets reference in /help, and you can hear a real Enhanced before/after in the homepage demos. Below is what each approach does for your audio, its trade-offs, and the decision tree we would use ourselves.

Two model families, two ways of working

Enhanced uses a trained audio model and GPU processing to synthesize high-frequency detail, aiming for greater definition and air. This is neural reconstruction, not simply turning up the treble or smoothing noise. The model generates plausible detail from your source; it cannot recover the exact information that was lost. Texture and timbre can change, so listen to the result alongside the original.

Flow Matching uses a different generative restoration approach. Guided by the source, it rebuilds detail through repeated refinement. That iterative work requires substantially more compute than a simple filter and can reshape texture, transients and ambience. Normal and High use different processing depths; High uses more compute without guaranteeing a better result.

The distinction matters because the two approaches can produce different textures and can fail in different ways. Flow Matching may invent plausible structure that was not present in the source, while Enhanced follows a shorter reconstruction path. That makes Flow Matching an alternative to audition, not an automatic upgrade.

What each preset is good at

Enhanced shines on lossy-codec damage. The 11 kHz rolloff on a 128 kbps mp3, quantization noise, smeared transients from low-bitrate encoders — that whole class of artifact is what its training was built around. On a Suno mp3 download or a YouTube-rip WAV, Enhanced typically delivers most of the perceptual improvement you would get from any preset through the shortest processing path, for one credit.

Flow Matching offers a different, iterative reconstruction. Its resynthesis can change cymbals, breath noise, reverb tails, and transient texture, but whether that change is an improvement depends on the source and the listener. It should be judged by A/B listening, not assumed to sit above Enhanced.

Enhanced exports WAV at 44.1 kHz and 24-bit by default, with 48 kHz on Pro and Studio. Flow Matching historically added a 32-bit float option, but that preset and option are temporarily unavailable.

Where each one falls short

Enhanced is the shorter, lower-credit option, but no refinement model suits every source. Its reconstruction can sound too uniform or leave characteristics a listener dislikes on some tracks. That is a reason to audition an alternative, not evidence that Flow Matching is a higher-quality tier.

Flow Matching was slower and more expensive. Six to twelve minutes per three-minute track at 3-4 credits was real cost when running through an album's worth of material. Like Enhanced, it generates a reconstruction rather than recovering an original master. It can add detail you did not intend or alter the character of a sound. Keep your original and compare at matched listening levels; neither preset is a guaranteed improvement on every recording.

The time, cost, and output trade-off

Enhanced costs 1 credit and is the shortest processing path, but its end-to-end time still varies with track length, queue demand, and worker startup. Before the pause, Flow Matching took roughly 6-12 minutes on a three-minute track and cost 3 credits on Normal or 4 on High — plus a cleaning stage that could take roughly the track's length when AI artifact removal was enabled. The higher runtime and credit cost were the headline trade-off.

There is no consistent quality ranking between the two presets. Some tracks sound better to a listener through Enhanced, some through Flow Matching, and sometimes the difference is not useful. Start with Enhanced. If Flow Matching returns, only spend the extra time and credits when you explicitly want to compare an alternative output.

AI-detector behavior is not a preset feature

Neither preset targets, defeats, or guarantees a result from an AI-content detector. Classifiers change, disagree, and can make false calls. Detector behavior should not be used to market Flow Matching as an upgrade over Enhanced.

Choose by listening and keep the source AI provenance disclosed wherever a platform or law requires it. Processing does not change that responsibility.

For the full release-prep workflow including loudness, format, and platform-specific notes, the longer write-up is in how to master Suno tracks for release.

A decision tree

Default to Enhanced. On most material — Suno mp3 downloads, YouTube rips, lossy renders from any source — it does enough of the job that the extra cost and wait of Flow Matching are not justified. Run it first, listen on whatever you would normally mix on, and ship it if it sounds right.

Flow Matching is currently unavailable, so there is no second preset to audition on new work. Existing results can still be downloaded and compared, but use Enhanced for any new refinement.

If you are unsure, run Enhanced first. It costs one credit and is the shortest processing path. If the output sounds finished, you are done. If it does not, keep the source and contact support with the job ID so the result can be investigated. Do not spend more credits trying to work around a quality problem.

Plans and credits

Plan limits and credit costs are on the pricing page. Briefly: Free is 2/month, Hobby covers light use, Pro and Studio unlock 48 kHz output, and one-off credit packs cover surge use without changing your plan. Join the waitlist if you have not already — use Enhanced while Flow Matching remains paused.