LFM3-Audio-2B is a compact, fully featured audio model that runs locally on laptops and edge devices. It takes text or audio as input and responds with text, speech, or both.
Alternatives to LFM3-Audio-2B
114 open-weight models that do the same job as LFM3-Audio-2B (2B omni), ranked by how close they are in task and size, then by popularity and how recent they are. Licence and VRAM are shown for each.
Compare LFM3-Audio-2B with the top 3LFM3-Omni-2B is Liquid AI's first omni-modal model in the LFM3 family: a single end-to-end model that understands text, images, video and audio, and answers in text and speech. It is built for on-device use: one model replaces a separate ASR model, vision-language model, LLM and TTS model.
2.2B125K ctx lfm1.027.3K 731Oct 2026- Same category
- Similar size (2.2B vs 2B)
- Newer release
d1-Omni-2B is a decision model for text, pictures, and audio, built on LFM3-Omni-2B. It returns calibrated answers in one forward pass, without generating tokens.
2.2B125K ctx lfm1.06.5K 127Oct 2026- Same category
- Similar size (2.2B vs 2B)
- Newer release
Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants.
5.1B128K ctx Apache 2.02.7M 1.2KMar 2026- Same category
- Larger (5.1B vs 2B)
- Commercial use OK
- Widely used
Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants.
8B128K ctx Apache 2.04.3M 1.9KMar 2026- Same category
- Larger (8B vs 2B)
- Commercial use OK
- Widely used
Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants.
5.1B128K ctx Apache 2.093.3K 569Mar 2026- Same category
- Larger (5.1B vs 2B)
- Commercial use OK
Janus is a novel autoregressive framework that unifies multimodal understanding and generation. It addresses the limitations of previous approaches by decoupling visual encoding into separate pathways, while still utilizing a single, unified transformer architecture for processing.
2.1B16K ctx MIT1.3K 611Oct 2024- Same category
- Similar size (2.1B vs 2B)
- Commercial use OK
We present JanusFlow, a powerful framework that unifies image understanding and generation in a single model. JanusFlow introduces a minimalist architecture that integrates autoregressive language models with rectified flow, a state-of-the-art method in generative modeling.
2B16K ctx MIT748 171Nov 2024- Same category
- Similar size (2B vs 2B)
- Commercial use OK
This model is Rex-Omni, a 3B-parameter Multimodal Large Language Model (MLLM) presented in the paper "Detect Anything via Next Point Prediction". It is compatible with the Hugging Face transformers library and is licensed under the IDEA License 1.0.
4.1B125K ctx Custom licence14.7K 66Oct 2025- Same category
- Larger (4.1B vs 2B)
Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants.
8B128K ctx Apache 2.0423.5K 521Mar 2026- Same category
- Larger (8B vs 2B)
- Commercial use OK
Aero-1-Audio is a compact audio model adept at various audio tasks, including speech recognition, audio understanding, and following audio instructions.
2.4B32K ctx MIT185 91Apr 2025- Same category
- Similar size (2.4B vs 2B)
- Commercial use OK
Youtu-Parsing-Omni is a compact (5B) omni-modal parsing model. Given a single input — a document page, a natural image, a chart / flowchart, a geometry figure, an audio clip or an audio-visual video — it produces one structured JSON envelope that covers both perception (layout elements, text…
5.3B1M ctx youtu-parsing2.5K 132Oct 2026- Same category
- Larger (5.3B vs 2B)
- Newer release
Qwen2.5-Omni is an end-to-end multimodal model designed to perceive diverse modalities, including text, images, audio, and video, while simultaneously generating text and natural speech responses in a streaming manner.
5.5B32K ctx Qwen Research License98.4K 381Apr 2025- Same category
- Larger (5.5B vs 2B)
Skywork-UniPic is a unified autoregressive multimodal model with 1.5 billion parameters, capable of handling three key vision-language tasks within a single architecture:
1.5BMIT111 116Jul 2025- Same category
- Similar size (1.5B vs 2B)
- Commercial use OK
OmniNeural is the first fully multimodal model designed specifically for Neural Processing Units (NPUs). It natively understands text, images, and audio, and runs across PCs, mobile devices, automobile, IoT, and robotics.
4BCC BY 4.02.2K 169Aug 2025- Same category
- Similar size (4B vs 2B)
- Commercial use OK
Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants.
470M256K ctx Apache 2.0177.5K 347Apr 2026- Same category
- Smaller (470M vs 2B)
- Commercial use OK
d1-omni-600M is a 600M parameter decision model built on LFM2.5-Encoder-350M. You give it a state (text or JSON, with images or a voice clip) and a set of named questions.
587M125K ctx lfm1.08.4K 167Oct 2026- Same category
- Smaller (587M vs 2B)
- Newer release
OmniGen2 is a powerful and efficient unified multimodal model. Unlike OmniGen v1, OmniGen2 features two distinct decoding pathways for text and image modalities, utilizing unshared parameters and a decoupled image tokenizer. OmniGen2 has competitive performance across four primary capabilities:
4BApache 2.02.3K 462Jun 2025- Same category
- Similar size (4B vs 2B)
- Commercial use OK
Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants.
420M256K ctx Apache 2.0335.3K 205Apr 2026- Same category
- Smaller (420M vs 2B)
- Commercial use OK
MOSS-Audio is an open-source audio understanding model from MOSI.AI, the OpenMOSS team, and Shanghai Innovation Institute.
5.2B40K ctx Apache 2.02.4K 91Apr 2026- Same category
- Larger (5.2B vs 2B)
- Commercial use OK
Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models. Gemma 3n models are designed for efficient execution on low-resource devices.
5.4B32K ctx Gemma Terms of Use25.2K 336Jun 2025- Same category
- Larger (5.4B vs 2B)
Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants.
12B256K ctx Apache 2.01.6M 1.8KMay 2026- Same category
- Larger (12B vs 2B)
- Commercial use OK
- Widely used
SEA-LION is a collection of Large Language Models (LLMs) which have been pretrained and instruct-tuned for the Southeast Asia (SEA) region.
5.1B128K ctx MIT1.2K 19May 2026- Same category
- Larger (5.1B vs 2B)
- Commercial use OK
Janus-Pro is a novel autoregressive framework that unifies multimodal understanding and generation. It addresses the limitations of previous approaches by decoupling visual encoding into separate pathways, while still utilizing a single, unified transformer architecture for processing.
1B16K ctx MIT3.1K 507Jan 2025- Same category
- Smaller (1B vs 2B)
- Commercial use OK
Music Flamingo (MF) is a fully open, state-of-the-art Large Audio-Language Model (LALM) designed to advance music (including song) understanding in foundational audio models. MF brings together innovations in:
8.3B32K ctx Custom licence94.2K 116Nov 2025- Same category
- Larger (8.3B vs 2B)