LongCat-2.5 is a native multimodal MoE model for long-horizon agentic tasks, with 1.6 trillion total parameters and approximately 48 billion activated per token.
Alternatives to LongCat-2.5
120 open-weight models that do the same job as LongCat-2.5 (1.78T vision), 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.
Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.
873M256K ctx Apache 2.0867.8K 141Feb 2026- Same category
- Smaller (873M vs 1.78T)
- Commercial use OK
apex-flash-1-abliterated is an experimental derivative of apex-flash-1, Cantina Security's open-weights security model developed in partnership with Yeta (@yetalabs on X).
321BA18.3B1M ctx MIT710 51Sept 2026- Same category
- Smaller (321B vs 1.78T)
- Commercial use OK
Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.
125BA10B256K ctx Apache 2.0767.6K 658Feb 2026- Same category
- Smaller (125B vs 1.78T)
- Commercial use OK
This is an uncensored version of Qwen/Qwen3.8-27B created with abliteration (see remove-refusals-with-transformers to know more about it). This is a crude, proof-of-concept implementation to remove refusals from an LLM model without using TransformerLens.
27.8B256K ctx Apache 2.0140.7K 800Aug 2026- Same category
- Smaller (27.8B vs 1.78T)
- Commercial use OK
Muse Glimmer is a 30-billion-parameter causal language model with a dedicated perception encoder, distilled from Muse Spark and purpose-built for autonomous agentic tasks on consumer hardware.
29.8B128K ctx Apache 2.0155.7K 2.2KAug 2026- Same category
- Smaller (29.8B vs 1.78T)
- Commercial use OK
Muse Glimmer is a 30-billion-parameter causal language model with a dedicated perception encoder, distilled from Muse Spark and purpose-built for autonomous agentic tasks on consumer hardware.
2.6B128K ctx Apache 2.0149.4K 83Aug 2026- Same category
- Smaller (2.6B vs 1.78T)
- Commercial use OK
Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.
4.7B256K ctx Apache 2.0503.2K 133Feb 2026- Same category
- Smaller (4.7B vs 1.78T)
- Commercial use OK
ZAYA1-VL-8B is a vision-language model (VLM) built upon Zyphra's ZAYA1-8B LLM. It has state-of-the-art performance among VLMs for its size and inference efficiency.
9.7B32K ctx Apache 2.0338.6K 43Apr 2026- Same category
- Smaller (9.7B vs 1.78T)
- Commercial use OK
The advanced capabilities of the ERNIE 4.5 models, particularly the MoE-based A47B and A3B series, are underpinned by several key technical innovations:
424BA47B128K ctx Apache 2.0973 107Jun 2025- Same category
- Smaller (424B vs 1.78T)
- Commercial use OK
Clef is a 27B multimodal model that turns a state and a schema of typed questions into decisions. It reads the state as text, JSON, images, or video, and returns a probability for every allowed option of every question in a single forward pass.
27.4B256K ctx Apache 2.078K 2.4KSept 2026- Same category
- Smaller (27.4B vs 1.78T)
- Commercial use OK
Interfaze 1.1 Lite is a mixture-of-architectures (MoA) model for developer workloads: reading documents, transcribing speech, locating objects and interface elements, and answering over all of it with structured output.
?Binterfaze-open146 69Oct 2026- Same category
Meet Qwen3-VL — the most powerful vision-language model in the Qwen series to date.
2.1B256K ctx Apache 2.01.1M 521Oct 2025- Same category
- Smaller (2.1B vs 1.78T)
- Commercial use OK
NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16 is the full-precision release of NVIDIA Nemotron 3.5 Super VL, a vision-language model for reasoning, coding, tool use, document understanding, and video understanding.
124BA12B1M ctx openmdw-1.174.6K 131Oct 2026- Same category
- Smaller (124B vs 1.78T)
- Commercial use OK
FLUX 3 Image Klein Prompt Upsampler is a 9B vision-language model that expands a short image request into the detailed prompt FLUX 3 Image Klein works best with. It takes the request, up to 10 reference images and an optional list of bounding boxes, and returns one prompt.
9.4B256K ctx Apache 2.064.7K 82Oct 2026- Same category
- Smaller (9.4B vs 1.78T)
- Commercial use OK
Meet Qwen3-VL — the most powerful vision-language model in the Qwen series to date.
236BA22B256K ctx Apache 2.081K 421Sept 2025- Same category
- Smaller (236B vs 1.78T)
- Commercial use OK
Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.
2.3B256K ctx Apache 2.0327.3K 108Feb 2026- Same category
- Smaller (2.3B vs 1.78T)
- Commercial use OK
NVIDIA Cosmos Reason 2 is an open, customizable, 2B-parameter reasoning vision language model (VLM) for physical AI and robotics that enables robots and vision AI agents to reason like humans, using prior knowledge, physics understanding and common sense to understand and act in the real world.
2.4B256K ctx NVIDIA Open Model License496.1K 413Dec 2025- Same category
- Smaller (2.4B vs 1.78T)
- Commercial use OK
VLX-Seek-1.5-10B is the open-source 10B model in the VLX-Seek 1.5 family, designed for fine-grained perception and visual grounding in embodied scenarios.
10B256K ctx Apache 2.088.5K 41Jul 2026- Same category
- Smaller (10B vs 1.78T)
- Commercial use OK
A Pocket-Sized MLLM for Ultra-Efficient Image and Video Understanding on Your Phone
1.3B256K ctx Apache 2.0188K 1.3KApr 2026- Same category
- Smaller (1.3B vs 1.78T)
- Commercial use OK
Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.
9.7B256K ctx Apache 2.0265K 159Feb 2026- Same category
- Smaller (9.7B vs 1.78T)
- Commercial use OK
In the past five months since Qwen2-VL’s release, numerous developers have built new models on the Qwen2-VL vision-language models, providing us with valuable feedback. During this period, we focused on building more useful vision-language models.
8.3B125K ctx Apache 2.05.4M 1.8KJan 2025- Same category
- Smaller (8.3B vs 1.78T)
- Commercial use OK
- Widely used
WeVisOne-VL (WeChat Vision All-in-One) is a unified vision-language model developed by WeChat Vision, Tencent Inc. Built upon Qwen3-VL, it unifies object-level perception (open-vocabulary object detection, referring localization, object counting, object pointing, and OCR) and pixel-level perception…
4.8B256K ctx Apache 2.034.9K 128Oct 2026- Same category
- Smaller (4.8B vs 1.78T)
- Commercial use OK
The advanced capabilities of the ERNIE 4.5 models, particularly the MoE-based A47B and A3B series, are underpinned by several key technical innovations:
424BA47B128K ctx Apache 2.0457 91Jun 2025- Same category
- Smaller (424B vs 1.78T)
- Commercial use OK
Clef-Flash is a 9B multimodal model that turns a state and a schema of typed questions into decisions. It reads the state as text, JSON, images, or video, and returns a probability for every allowed option of every question in a single forward pass.
9.4B256K ctx Apache 2.033.9K 929Sept 2026- Same category
- Smaller (9.4B vs 1.78T)
- Commercial use OK