LLMDB
Contribute
H Company logo

Holo2 235B-A22B Preview

H CompanyOpen WeightsPending Human Review

Holo2 235B-A22B Preview is H Company’s separately published large research tier for visual interface localization. Built on Qwen3-VL-235B-A22B-Thinking, it uses sparse language computation with about 22B active parameters per token. The model interprets a textual request and high-resolution screenshot to identify the relevant interface element, and can refine localization through repeated agent steps. The February 3 launch confirms downloadable weights; repository creation in January is not used as publication. It is a language-capable vision model rather than an embedding or detection-only endpoint, but its published evaluation scope centers on grounding instead of broad autonomous-agent performance.

2026-02-03
235B language tier, 22B active
Qwen3-VL Mixture-of-Experts vision-language Transformer
CC-BY-NC-4.0
Compare with other models

Specifications

Parameters
235B language tier, 22B active
Architecture
Qwen3-VL Mixture-of-Experts vision-language Transformer
License
CC-BY-NC-4.0
Type
multimodal
Modalities
textimage

Benchmark Scores

Advanced Specifications

Model Family
Holo
Finetuned From
Qwen3-VL-235B-A22B-Thinking
API Access
Not Available
Chat Interface
Not Available
Variants
Holo2-235B-A22B

Capabilities & Limitations

Capabilities
high-resolution UI localizationvisual groundingiterative refinement
Known Limitations
Research preview primarily evaluated for localizationNon-commercial metadata overrides conflicting generic Apache card proseFull weights require substantial serving memory
Notable Use Cases
GUI grounding researchcomputer-use perceptionhigh-resolution screen analysis

Resources

Related Models

Holo4 27B

H Company

Holo4 27B

Holo4 27B is H Company’s computer-use vision-language model built on Qwen3.8-27B. It is post-trained to work across graphical interfaces, code, MCP tools, and APIs, using screenshots and tool results to choose its next actions. Released weights support local agent research; an external harness executes clicks, typing, and code and returns observations. The dense 27B model uses a non-commercial weight license, unlike the Apache-licensed sparse Holo4 tier. Its card specifies a 262,144-token configuration limit; this is not a promise of consistent task performance throughout that window.

Typemultimodal
Parameters27B
2026-09-28
Open Weights
DetailsCompare

Holo4 35B-A3B

H Company

Holo4 35B-A3B

Holo4 35B-A3B is H Company’s computer-use vision-language model built on Qwen3.6-35B-A3B. It is post-trained to work across graphical interfaces, code, MCP tools, and APIs, using screenshots and tool results to choose its next actions. Released weights support local agent research; an external harness executes clicks, typing, and code and returns observations. The sparse model activates about 3B language parameters per token but still needs memory for its full checkpoint. Its card specifies a 262,144-token context. Apache-licensed weights and hosted H Models API access are available.

Typemultimodal
Parameters35B total, 3B active
2026-09-28
Open Weights
DetailsCompare

Holotron4 Nano

H Company

Holotron4 Nano

Holotron4 Nano is H Company’s computer-use vision-language model built on NVIDIA Nemotron 3 Nano Omni. It is post-trained to work across graphical interfaces, code, MCP tools, and APIs, using screenshots and tool results to choose its next actions. Released weights support local agent research; an external harness executes clicks, typing, and code and returns observations. The sparse model activates about 3B language parameters per token but still needs memory for its full checkpoint. Its card specifies a 262,144-token context. The NVIDIA Open Model Agreement governs its weights; the released page establishes text and image understanding without assuming every modality of its Omni backbone.

Typemultimodal
Parameters30B language tier, 3B active
2026-09-28
Open Weights
DetailsCompare