Open Model WeightsLIQUID AI · FIELD VERIFIED
Primary sources last checked on 4 Nov 2026. Verification freshness is tracked separately from factual change, and unknown values remain explicit rather than inferred.
CONCEPTUAL MODEL BRIEF · ORIGINAL ILLUSTRATION
A topic-specific editorial visual, not an independent model performance measurement.MODEL INTELLIGENCE · SOURCE-FIRST
Open Model Weights records LFM3-Omni-2B as a 2.16693B-parameter text, images, video, audio → text and speech open-weight model in the LFM3 Omni family, with Liquid AI listed as the developer. The checked source repository provides 2 recognized weight artifacts, format evidence for Safetensors and a verified context window of 128,000 tokens. Values that are not evidenced in the checked sources remain explicitly undisclosed or unknown.
The repository declares other. Commercial-use classification requires license-specific review. Open Model Weights reports the checked license evidence and classification; this is not legal advice.
No base model is declared in the checked standard lineage metadata, so Open Model Weights does not infer one. Training recipe status: training-described-in-model-card; data disclosure status: not-disclosed-in-standard-metadata-or-obvious-model-card-section.
+ Save to my shortlistMonitor changes →Local browser monitoring · no background email alerts
DIRECT ANSWER · FEATURED MODEL BRIEFING
Source-grounded editorial layer · reviewed 2026-11-04
Liquid AI LFM3-Omni-2B combines text, images, audio and sampled video inputs with text or generated speech outputs. Public model and codec weights are available under the custom LFM Open License; code and modality support need careful verification. Open Model Weights independently inspected the exact repository revision, listed weight artifacts and declared license; it has not run the model or reproduced publisher benchmark results. The repository includes a separately checked license file.
WHY IT MATTERS
A 2.17B multimodal checkpoint that can take image, video, audio and text input and respond in text or speech. The exact configuration and operational assumptions should be checked against the source model card rather than assumed from the name alone.
LFM Open License v1.0 is a custom rights framework; repository license file exists but commercial-use status is not resolved automatically. The exact configuration and operational assumptions should be checked against the source model card rather than assumed from the name alone.
Liquid AI published omni, audio and vision benchmark tables; OMW has not reproduced these measurements. The exact configuration and operational assumptions should be checked against the source model card rather than assumed from the name alone.
MODEL-SELECTION VIEW
A 2.17B multimodal checkpoint that can take image, video, audio and text input and respond in text or speech. The exact configuration and operational assumptions should be checked against the source model card rather than assumed from the name alone.
The repository LiquidAI/LFM3-Omni-2B was checked at commit 21d75123f86c1f7ffb4bef1ea202710c4e7ebe67. This evidence establishes a real published weight distribution, but no independent user task or hardware benchmark was performed by Open Model Weights.
Its public Safetensors include the main checkpoint and audio tokenizer components for a combined multimodal pipeline. The exact configuration and operational assumptions should be checked against the source model card rather than assumed from the name alone.
The published weight manifest and associated processor or tokenizer components must be pinned to 21d75123f86c1f7ffb4bef1ea202710c4e7ebe67 to avoid silently mixing revisions. Converted quantizations and separately licensed dependencies are outside an automatic equivalence guarantee.
LFM Open License v1.0 is a custom rights framework; repository license file exists but commercial-use status is not resolved automatically. The exact configuration and operational assumptions should be checked against the source model card rather than assumed from the name alone.
The exact repository license identifier is other. Commercial-use clearance is recorded only as supported by checked source metadata and text; third-party assets, deployment software, encoders and data sources retain their own terms.
Liquid AI published omni, audio and vision benchmark tables; OMW has not reproduced these measurements. The exact configuration and operational assumptions should be checked against the source model card rather than assumed from the name alone.
Publisher benchmarks are useful context but depend on their chosen prompts, hardware, code revision and comparison conditions. Open Model Weights does not claim to have reproduced any reported ranking or speed figure.
Transformers reference uses trust_remote_code and a speech codec; not every backend supports every output mode. The exact configuration and operational assumptions should be checked against the source model card rather than assumed from the name alone.
Actual deployment of Liquid AI LFM3-Omni-2B requires consistent model code, configuration, runtime, security permissions and memory budget. Do not infer general usability solely from public downloads or parameter count.
IMPLEMENTATION & EVALUATION
Reproducible operator checklist; these steps are not claims of an OMW model run.
A named model does not guarantee identical checkpoint revisions.
Record 40-character SHA, original file names, configuration and tokenizer/processor versions.
Unpinned revisions can silently change weights or usage requirements.
A specialist checkpoint can fail outside the publisher's demonstrated conditions.
Document actual prompt templates, modality preprocessing, inference backend and compatible software versions.
Using unsupported serving backends can invalidate comparisons.
Model card labels and code licenses are not automatically permission for every deployed use.
Check exact weight terms, notices, tool permissions, data privacy and runtime isolation.
Restricted weight rights or unsafe action tools can make a deployment unsuitable.
Vendor benchmark figures are useful evidence, but are not OMW-run observations.
Pre-register held-out tasks and compare baseline checkpoint, hardware, quality, memory and error rates.
Attributing first-party results to independent inference would mislead readers.
Proposed tests, not OMW benchmark results. Read the official model card at the pinned source revision ↗ Inspect OMW field evidence ↗
PRIMARY & INDEPENDENT SOURCES
FAQ
A 2.17B multimodal checkpoint that can take image, video, audio and text input and respond in text or speech. The official checkpoint must be distinguished from third-party conversions and derivatives.
Yes. Exact public Safetensors were found in LiquidAI/LFM3-Omni-2B at the pinned source revision. External dependencies may have different access conditions.
LFM Open License v1.0 is a custom rights framework; repository license file exists but commercial-use status is not resolved automatically. The source verifier only records evidenced permissions and does not issue legal clearance.
Liquid AI published omni, audio and vision benchmark tables; OMW has not reproduced these measurements. OMW has not run a separate evaluation.
Transformers reference uses trust_remote_code and a speech codec; not every backend supports every output mode. Operators should verify compatible software and isolate external tools.
Test synchronized video-audio questions, speech output latency, hallucinations and language coverage with pinned settings. No unperformed independent speed or quality measurement is claimed.
03 · PRACTICAL TECHNICAL PROFILE
Inspect weight-memory estimates, model lineage and runtime evidence separately from promotional or untested performance claims.
DEPLOYMENT PROFILE
These bars visualize weight-only decimal-GB estimates. They exclude KV cache, activations, runtime overhead, optimizer state and sharding.
LICENSE & USE
Commercial-use classification requires license-specific review
This section summarizes the checked repository evidence. It is not legal advice. Inspect license evidence ↗
LINEAGE & PROVENANCE
Open Model Weights does not infer a hidden base model or training corpus when the checked sources do not declare one.
RUNTIME SIGNALS
Declared and model-card-mentioned runtimes are discovery signals. They are not treated as Open Model Weights runtime tests unless explicitly measured.
EVIDENCE TRAIL
This is where the article remains auditable: source links on one side, the registry's observed change history on the other.
Official Hugging Face repository SHA unchanged; prior field evidence retained.
Checked Hugging Face API metadata, exact repository file list, model card and config when accessible.
04 · EVIDENCE AND UPDATES
Inspect dated source checks and any differences between published claims and the evidence; these checks are not independent model benchmarks.
FRESHNESS & VERIFICATION
A verification check does not pretend the model changed. Open Model Weights updates the verification date after a real primary-source check, while factual-update dates move only when retained source evidence changes.
CONFLICT INTELLIGENCE
05 · SOURCE RECORD
Full repository-level attributes, licensing, format details and provenance are available below without hiding or rewriting the original data.
01 · WEIGHTS
02 · LICENSE
Classification describes the checked license terms; it is not legal advice.
License / model-card evidence ↗03 · HARDWARE
Weight-only decimal-GB estimate from total parameter count; excludes KV cache, activations, runtime overhead, optimizer state and sharding.
04 · FORMATS
“Not observed” means no matching repository artifact/name/dtype signal was found; it does not prove no third-party conversion exists.
05 · LINEAGE
06 · TRAINING ASSETS
07 · RUNTIME SUPPORT
Open Model Weights separates model-repository evidence from external runtime-release evidence. An external support statement does not change the official model repository's weight formats and does not imply an Open Model Weights inference benchmark.
08 · SOURCES
09 · VERIFICATION HISTORY
Official Hugging Face repository SHA unchanged; prior field evidence retained.
Checked Hugging Face API metadata, exact repository file list, model card and config when accessible.