L
LIQUID AI / LiquidAI/Liquid-Codec

LiquidCodec

Neural codec with about 74 million parameters that encodes speech into discrete tokens and decodes them into audio. Liquid AI presented it independently on November 12; it was already part of LFM3-Audio and LFM3-Omni. It does not convert text to speech on its own.

01 / SUMMARY

The essential

ORIENTATIONStreaming neural speech codec

Primary use stated or inferred cautiously from official documentation.

CONTEXT24 kHz mono speech; 12.5 Hz frames with twelve codes per frame

Maximum input capacity when published by the source.

EXITDiscrete codes and reconstructed speech; approximately 1.65 kbps

Documented maximum generation limit.

ACCESSHugging Face · Transformers

Verified availability channels.

MODEL TYPEAudio

Primary functional family and verified specialties.

LICENSELFM Open License v1.0 · conditional commercial use

Declared terms for API access, model weights, or self-hosting.

02 / TECHNICAL SHEET

Limits and integration

API IDLiquidAI/Liquid-Codec
Model typeAudio
Access modelDownloadable weights
LicenseLFM Open License v1.0 · conditional commercial use
DeploymentSelf-hosted
Release12 NOV 2026
Knowledge cutoffNot published
EntranceAudio · Audio tokens
ExitAudio tokens · Audio
Context window24 kHz mono speech; 12.5 Hz frames with twelve codes per frame
maximum outputDiscrete codes and reconstructed speech; approximately 1.65 kbps
ReasoningNot published
Published tools
Structured OutingsNot published
Batch processingNot published
Prompt cacheNot published
Fine-tuningNot published
Verified platformsHugging Face · Transformers
BEST SUITED FOR

Tokenize and reconstruct speech in local systems with streaming encoding and decoding. Its twelve codebook groups can be predicted in parallel. It accepts 24 kHz mono speech; it is neither a conversational assistant nor a transcriber.

WORTH MONITORING

Evaluations cover speech and do not establish quality for music, singing or environmental sounds. The BF16 repository includes custom code: review the implementation and pin a revision before loading it. Quantized decoders are integrated into LFM3 GGUF distributions, not separate checkpoints in this repository.

03 / CAPABILITIES

What it can do

01

Orientation

Tokenize and reconstruct speech in local systems with streaming encoding and decoding. Its twelve codebook groups can be predicted in parallel. It accepts 24 kHz mono speech; it is neither a conversational assistant nor a transcriber.

02

Context

24 kHz mono speech; 12.5 Hz frames with twelve codes per frame · Discrete codes and reconstructed speech; approximately 1.65 kbps

03

Tools and integration

Not published

04

Access

Hugging Face · Transformers

Input modalities
AudioAudio tokens
04 / PRICES

Documented cost

ConceptWorthUnit/condition
Standard inputNot publishedStandard API rate
Cached inputNot publishedReading reused prefixes
Cache write or storageNot publishedThe condition varies by provider
Standard outputNot publishedMay include reasoning tokens
Batch inputNot publishedAsynchronous processing
Batch outputNot publishedAsynchronous processing

No hosted API tariff is verified; local deployment consumes compute. LFM Open License v1.0 conditions commercial use on an annual revenue threshold of USD 10 million. Review its terms; this is neither Apache 2.0 nor unconditional commercial permission.

i Prices change and may depend on level, region or context length. Check the source before making a decision.

05 / EVALUATIONS

How to read the results

“

Published speed measurements use FP32 and one Apple M4 Max CPU thread. Lower RTF values mean faster processing and are not full-application latency. ViSQOL 4.319 measures BF16 reconstruction on English LibriSpeech test-clean, not TTS generation or an average across languages.

BenchmarkResultMetricSource
Encoding RTF (Apple M4 Max, FP32)0,0028RTF · lower is betterView source ↗
Decoding RTF (Apple M4 Max, FP32)0,0115RTF · lower is betterView source ↗
ViSQOL (LibriSpeech test-clean, BF16)4,319ViSQOL · higher is betterView source ↗

Inferama only highlights a “best result” when the metric, test set, configuration, and date allow for an equivalent comparison. The supplier's figures are presented as claims from its own source.

06 / VERSIONS

Chronology

LiquidAI/Liquid-Codec

Standalone publication of the codec already used by LFM3-Audio and LFM3-Omni; repository creation is not the announcement date.

ANALYSIS

Related analysis

07 / SOURCES

Traceability