Hy4 flash is our most cost-effective Hy model to date, featuring 220B total parameters (including 64B Engram parameters) with just 12B active parameters per token. It natively accepts text and image inputs, generates text, and supports a context window of up to 1M tokens.
Alternatives to Hy4-flash
120 open-weight models that do the same job as Hy4-flash (220B 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.
UI-TARS is a next-generation native GUI agent model designed to interact seamlessly with graphical user interfaces (GUIs) using human-like perception, reasoning, and action capabilities.
73.4B32K ctx Apache 2.0252 25Jan 2025- Same category
- Smaller (73.4B vs 220B)
- Commercial use OK
We introduce InternVL2.5-MPO, an advanced multimodal large language model (MLLM) series that demonstrates superior overall performance. This series builds upon InternVL2.5 and Mixed Preference Optimization.
78.4B32K ctx Qwen License124 55Dec 2024- Same category
- Smaller (78.4B vs 220B)
Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.
4.7B256K ctx Apache 2.0496.7K 133Feb 2026- Same category
- Smaller (4.7B vs 220B)
- 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 2.0339.4K 43Apr 2026- Same category
- Smaller (9.7B vs 220B)
- Commercial use OK
Kimi K3.1 Pro is an open-weight, natively multimodal agentic model, and our most capable 3T-class model to date. Post-trained from Kimi K3 with large-scale joint reinforcement learning across hundreds of downstream domains, it is the first model in the K3.1 series.
2.78T1M ctx kimi-k3.1-pro68K 2.4KNov 2026- Same category
- Larger (2.78T vs 220B)
Meet Qwen3-VL — the most powerful vision-language model in the Qwen series to date.
2.1B256K ctx Apache 2.01.2M 522Oct 2025- Same category
- Smaller (2.1B vs 220B)
- 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.
26.5BA4B256K ctx Apache 2.062.3K 479Mar 2026- Same category
- Smaller (26.5B vs 220B)
- 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-open133 69Oct 2026- Same category
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 2.090.3K 83Aug 2026- Same category
- Smaller (2.6B vs 220B)
- 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 2.0321.5K 108Feb 2026- Same category
- Smaller (2.3B vs 220B)
- Commercial use OK
Inkling 2 Mini is a general-purpose multimodal model that accepts text and image inputs and generates text outputs. It is intended for use in English and other languages, and across multiple coding languages.
26.4BA3.5B1M ctx Apache 2.07.9K 3.1KNov 2026- Same category
- Smaller (26.4B vs 220B)
- Commercial use OK
We introduce Intern-S2-Mobius, a 35B foundation model built on the Mobius-v0 architecture realized by Xtuner and LMDeploy.
36B256K ctx Apache 2.01K 66Jul 2026- Same category
- Smaller (36B vs 220B)
- Commercial use OK
AesCode generates information-rich visual artifacts such as slides, posters, and dashboards as HTML/CSS. The output remains structured, editable, and verifiable, but the task poses a distinct challenge: code models cannot see how layout, hierarchy, and color come together on the canvas.
33.4B256K ctx Apache 2.01.3K 58Sept 2026- Same category
- Smaller (33.4B vs 220B)
- Commercial use OK
We introduce Intern-S2-Preview, an efficient 35B scientific multimodal foundation model. Beyond conventional parameter and data scaling, Intern-S2-Preview explores task scaling: increasing the difficulty, diversity, and coverage of scientific tasks to further unlock model capabilities.
36.1BA4.9B256K ctx Apache 2.01.6K 123May 2026- Same category
- Smaller (36.1B vs 220B)
- Commercial use OK
A Pocket-Sized MLLM for Ultra-Efficient Image and Video Understanding on Your Phone
1.3B256K ctx Apache 2.0194.5K 1.3KApr 2026- Same category
- Smaller (1.3B vs 220B)
- 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 2.0266.8K 163Feb 2026- Same category
- Smaller (9.7B vs 220B)
- 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 License475.6K 420Dec 2025- Same category
- Smaller (2.4B vs 220B)
- Commercial use OK
One Gateway. Every Model. — Route Smarter · Ship Safer · Spend Less.
27.8BApache 2.08K 547Aug 2026- Same category
- Smaller (27.8B vs 220B)
- 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 2.05.2M 1.8KJan 2025- Same category
- Smaller (8.3B vs 220B)
- Commercial use OK
- Widely used
For NEO Imatrix GGUFS - reg and MTP, as well as full model details/card please go here:
40B256K ctx Apache 2.0337 68Jul 2026- Same category
- Smaller (40B vs 220B)
- 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 2.073.4K 42Jul 2026- Same category
- Smaller (10B vs 220B)
- Commercial use OK
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 2.036.4K 129Oct 2026- Same category
- Smaller (4.8B vs 220B)
- Commercial use OK
Flora-rewarder-27B produces outcome rewards (an episode-level success score), used for monitoring real-world deployments, and process rewards (dense action-quality scores over short time steps), used during reinforcement learning-based post-training.
27.8B256K ctx NVIDIA Non-Commercial License4.9K 23Oct 2026- Same category
- Smaller (27.8B vs 220B)
LFM2.5-VL-3B is a multimodal variant of LFM2.5, a family of hybrid models designed for on-device deployment. It builds on LFM2-VL-3B with further mid- and post-training.
3.1B32K ctx lfm1.054.2K 239Aug 2026- Same category
- Smaller (3.1B vs 220B)