Hugging Face · directory match
Qwen3.8-27B-NVFP4 is a 27-billion-parameter model hosted on Hugging Face and intended for image-text-to-text workloads. Its listed modalities are image and text, and its documented use case is vision, making it potentially relevant for applications that need to interpret or respond to visual inputs alongside written prompts. The available record does not disclose a context window, inference-speed benchmark, pricing, detailed capability tags, or task-specific evaluation results, so developers should validate accuracy, latency, memory requirements, and deployment compatibility independently. The model card content provided is limited and does not describe architecture, quantization behavior, recommended hardware, prompt format, safety characteristics, or production limitations. Licensing terms are not disclosed in the available source content; review the Hugging Face repository directly before redistribution or commercial use. The model is published under the PassingByPixels namespace on Hugging Face. The source page is currently indexed as an image-text-to-text model, but no publication, update, or release date is disclosed, so the freshness of its documentation and files cannot be established from the available material.
vision
- Size
- 27B parameters
- License tag
- Check source card
Verify model card ↗Hugging Face · directory match
elliot-vl-2b-v1.1-mid-train is listed as a 2-billion-parameter vision model that accepts text and image inputs, but its current documentation is unavailable. The model record identifies vision as its use case and lists text and image modalities, making it potentially relevant for applications that combine written prompts with visual content. However, the available Hugging Face source currently returns a 404 page rather than a model card, so architecture details, supported image formats, input limits, prompting guidance, implementation examples, evaluation results, and actual inference performance are not disclosed. No capabilities beyond the recorded modalities and vision use case are documented, and no tokens-per-second benchmark is reported. Licensing terms, safety guidance, training data, hardware requirements, and deployment instructions are also not disclosed in the available source. The recorded 2-billion parameter count is explicit in the model record, but practical suitability cannot be assessed confidently until the repository or an authoritative model card becomes available. Developers should treat this entry as a provisional index record, verify the repository path and model files independently, and avoid relying on undocumented behavior or compatibility assumptions.
vision
- Size
- 2B parameters
- License tag
- Check source card
Verify model card ↗Hugging Face · directory match
danAI-55M-Reasoning is a 55-million-parameter text-generation model listed on Hugging Face for agentic workflows, tool-calling, and reasoning-oriented use cases. The available model record identifies text as its only modality and lists agentic and tool-calling as capabilities, making it potentially relevant for lightweight agents or applications that need structured interaction with external tools. However, the supplied source content does not disclose context-window size, inference speed, benchmark results, pricing, training data, architecture, or deployment requirements, so practical performance and compatibility cannot be assessed from the available documentation. The model page is hosted at https://huggingface.co/asjadilahi/danAI-55M-Reasoning, and its current indexed description is only “text-generation”; detailed model-card information is not present in the supplied page capture. Licensing terms, safety behavior, limitations, and maintenance status are also not disclosed, so developers should inspect the live Hugging Face repository and its files before production use. The source was available for this review, but no publication or update date is provided, limiting confidence about documentation freshness.
agentstool-callingreasoning
- Size
- 55M parameters
- License tag
- Check source card
Verify model card ↗Hugging Face · directory match
Qwen3.6-35B-A3B-heretic-moe-max is listed as a 35-billion-parameter model with image and text modalities, and the directory record classifies it for uncensored, abliterated, and vision-oriented use cases. Those classifications may make it relevant for developers evaluating multimodal experimentation or models intended for fewer behavioral restrictions, but the available source does not provide technical documentation explaining its vision implementation, supported image formats, prompt format, deployment requirements, or quality characteristics. The model record does not disclose a context window, inference speed, pricing, or hosted API availability. Licensing terms are also not disclosed, so commercial and redistribution suitability should be verified before adoption. Most importantly, the referenced Hugging Face URL currently returns a 404 page rather than a model card or repository, meaning the model’s files, documentation, compatibility details, and current availability cannot be independently confirmed from the supplied source. Treat the directory metadata as provisional and verify the repository or an authoritative replacement before integrating it into production. No benchmark results or performance rankings are available.
vision
- Size
- 35B parameters
- License tag
- Check source card
Verify model card ↗Hugging Face · directory match
qwen3.8-9b-uncensored-cyber-exploit-XRPL-v3 is a 9-billion-parameter text model hosted on Hugging Face, with an uncensored classification indicated directly by its model name. Its listed modality is text, and the recorded use cases are agents, uncensored applications, and cybersecurity. This may make it relevant for developers evaluating text-based agent workflows or cybersecurity-oriented experimentation, but the supplied source does not document specific tools, function-calling support, coding performance, reasoning benchmarks, or deployment requirements. Context-window size, inference speed, pricing, and quantitative evaluation results are not disclosed. The supplied page snapshot identifies the Hugging Face repository and presents it as a text-generation model, but it does not expose a substantive model card, detailed methodology, training information, or safety documentation. Licensing terms are also not disclosed in the available source content, so review the repository’s current license and usage conditions before redistribution or commercial deployment. No dated release or update information is available, limiting confidence about source freshness; verify the live repository and files before relying on this listing for production decisions.
agents
- Size
- 9B parameters
- License tag
- Check source card
Verify model card ↗Hugging Face · directory match
Qwen3.8-27B-Uncensored-MTPLX-3bit is a 27-billion-parameter Hugging Face model intended for image-and-text-to-text use. Its listed modalities are image and text, and its documented classifications include unrestricted, uncensored, and abliterated; vision is therefore a relevant use case. A developer could evaluate it for multimodal prompting, image understanding, and applications requiring fewer restrictions, subject to independent validation of outputs and deployment behavior. The model name also identifies a 3-bit variant, but the supplied source does not document its runtime requirements, supported inference libraries, quantization format, or quality trade-offs. No context-window size, inference-speed benchmark, pricing, or performance evaluation is disclosed. The supplied page content exposes Hugging Face’s generic metadata and page title but does not provide a visible model card, detailed capabilities, examples, or benchmark results. Licensing terms are not disclosed in the available source, so review the repository’s licensing and usage information before redistribution or commercial deployment. The page is available at the supplied Hugging Face repository URL, but no publication or update date is exposed, so source freshness cannot be independently established from the provided material.
vision
- Size
- 27B parameters
- License tag
- Check source card
Verify model card ↗Hugging Face · directory match
RudolfV-2-S is a Hugging Face model indexed for image feature extraction and is most relevant to vision-oriented developer workflows. The available record identifies vision as its use case, but it does not disclose specific supported modalities beyond that classification, nor does it describe the model architecture, preprocessing requirements, output format, or supported tasks in detail. In practical terms, it may be worth evaluating for pipelines that derive features from images, such as visual search, similarity analysis, image organization, or downstream computer-vision systems, but those applications are not documented as guaranteed capabilities in the available source. No parameter count, context window, inference-speed benchmark, pricing, or deployment requirements are disclosed. The supplied Hugging Face page snapshot exposes the repository identity and canonical source URL, while the model card’s substantive technical documentation is not available in the provided content. Licensing terms, usage restrictions, training-data information, and safety guidance are also not disclosed, so review the repository directly before production or commercial use. The source snapshot contains no publication or update date, making its freshness impossible to verify; consult the live model card for current files, documentation, and license details.
vision
- Size
- Size not listed
- License tag
- Check source card
Verify model card ↗Hugging Face · directory match
Gemma-4-12B-OBLITERATED-cyber is a 12-billion-parameter image-text-to-text model hosted on Hugging Face. It accepts image and text inputs, making it potentially relevant for multimodal analysis and vision-oriented workflows. The model record identifies abliterated and cybersecurity as use-case classifications, so developers may consider it for cybersecurity-related experimentation and other applications involving image-text understanding. However, the available source content does not provide a model card description, context-window specification, inference-speed benchmark, pricing, evaluation results, usage instructions, or deployment requirements. The model name explicitly includes “OBLITERATED,” and the record classifies it as abliterated; this should not be treated as a broader safety or reliability guarantee. Licensing terms are not disclosed in the available source content, so review the Hugging Face repository and its files before redistribution or commercial use. The source page is available at Hugging Face, but the supplied content does not show a publication or last-update date, so source freshness cannot be independently established. Treat its capabilities and practical performance as insufficiently documented until the repository provides additional technical details.
vision
- Size
- 12B parameters
- License tag
- Check source card
Verify model card ↗Hugging Face · directory match
VideoGuard-Qwen3.5-4B-Safety-RL-Uncensored is a Hugging Face model intended for video-text-to-text workflows, with the directory record identifying vision, video, reasoning, and uncensored use cases. Its listed modalities are image and text, so developers should verify the repository documentation before assuming direct video-input support or a particular video-processing interface. The model is recorded as having 4 billion parameters, but the supplied page content does not expose technical details such as context length, inference speed, recommended hardware, quantization options, or evaluation results. Pricing is not disclosed because this is a Hugging Face repository rather than a hosted API listing. The model name explicitly includes “Uncensored,” supporting the uncensored classification, but that label should not be treated as a guarantee of behavior, reliability, or suitability for sensitive deployments. Licensing terms and usage restrictions are not disclosed in the supplied source and should be checked on the repository before redistribution or commercial use. The Hugging Face page is the current source referenced here; its visible metadata is primarily generic site content, so the model card appears insufficiently documented for production decisions without further repository review.
visionreasoning
- Size
- 4B parameters
- License tag
- Check source card
Verify model card ↗Hugging Face · directory match
chandra-ocr-2-BF16-GGUF-RDNA4 is a Hugging Face model repository intended for vision-oriented workloads involving image and text inputs. The available record classifies it under vision and identifies its modalities as image and text, making it potentially relevant for OCR or document-understanding experiments suggested by its name. However, the supplied source page does not disclose a model card with architecture details, supported tasks, context window, parameter count, evaluation results, inference speed, hardware requirements, or pricing. Developers should therefore avoid assuming a particular OCR accuracy level, throughput, deployment method, or compatibility beyond what is indicated by the repository name and modality metadata. The name indicates a BF16 GGUF variant associated with RDNA4, but the source does not provide validated runtime instructions or benchmarks for those characteristics. Licensing terms and usage restrictions are not disclosed in the supplied page content, so they should be verified directly in the repository before redistribution or commercial use. No publication or update date is available in the provided source, so source freshness cannot be independently assessed.
vision
- Size
- Size not listed
- License tag
- Check source card
Verify model card ↗Hugging Face · directory match
Qwen3.8-27B-Uncensored-JoyFox-Aggressive-GGUF is a 27-billion-parameter GGUF model hosted on Hugging Face, intended for local or compatible GGUF inference workflows. Its name explicitly identifies it as uncensored and aggressive, while the directory record classifies it as abliterated and lists vision as a use case. However, the supplied source page does not expose a substantive model card in the available content, so the exact vision implementation, supported image formats, prompt format, quantization variants, hardware requirements, context window, and expected performance are not disclosed. No inference speed benchmark or pricing information is provided. The source also does not disclose a license, training data, evaluation results, safety policy, or maintenance status; these should be checked on the Hugging Face repository before deployment. The record lists no modalities despite the vision classification, so modality support should be treated as requiring verification rather than assumed. This model may be relevant for experimentation with less-restricted text generation and possible multimodal use, but developers should validate behavior, compatibility, and content safeguards in their own environment. The source is the Hugging Face repository, and its available page content is current only as captured here; no freshness date is disclosed.
vision
- Size
- 27B parameters
- License tag
- Check source card
Verify model card ↗Hugging Face · directory match
0glm is a Hugging Face text-generation model intended for coding-related use cases. The available source identifies the repository at 0penAGI/0glm and presents it as a model page, but the supplied page content does not expose a model card with technical specifications, examples, evaluation results, or detailed capability descriptions. Its supported modality is text, so developers should not assume native image, audio, or video handling. The parameter count, context window, inference speed, pricing, and deployment requirements are not disclosed in the available source. Licensing terms and any restrictions are also not disclosed, so review the repository and its files before using the model commercially or redistributing it. Because implementation guidance and benchmarks are absent, suitability for production coding, agentic workflows, or tool-calling cannot be established from this listing alone. The page content provides no visible publication or update date, so source freshness is not independently verifiable; consult the live Hugging Face repository for current files, metadata, and documentation.
coding
- Size
- Size not listed
- License tag
- Check source card
Verify model card ↗