Search FoxAI

AI Models 84999

• synthetic-image-forensics↗

  broken-ice

  The Synthetic Image Forensics tool is used for synthetic image classification, defect localization, and explanation generation.

• nllb-200-distilled-600M-onnx↗

  subtitleking

  Meta's NLLB-200 distilled 600M translation model as fp16 ONNX is used for inference in the browser.

• smollm2-135m-tictactoe↗

  Sexyparabola

  A small chat model that can play tic-tac-toe. It has been LoRA fine-tuned and merged back into the base weights, so it can be loaded as a normal transformers model without PEFT dependencies.

• Llama-2-7b-chat-guanaco-GGUF↗

  laphilemon

  Llama-2-7b-chat-guanaco-GGUF is an instruction fine-tuned version of the Llama-2-7b-chat-hf model, quantized and trained using QLoRA technology.

• stock-picker-mcp-source↗

  surendracolorado

  Stock Picker MCP Server: System Architecture and Management Guide, provides a multi-agent automated swing trading research process, dynamically screening stocks, analyzing historical prices, and evaluating technical setups through MCP tool calls.

• tmp-ocr-freeform↗

  Hoffmann0412

  This tool is an experimental research repository for OCR Freeform, emphasizing the need for further testing, rather than generating scores or making statements.

• yama-indic↗

  LatentAILabs

  Yama is a zero-shot text classifier for Indian languages and Hinglish, using the XLM-RoBERTa-large encoder, capable of performing classification without fine-tuning.

• Qwen3.8-27B-W4A16-HGPTQ-g64↗

  nchapman

  Qwen3.8-27B-W4A16-HGPTQ-g64 is a quantized version of the Qwen/Qwen3.8-27B model, using the All-W4A16 Cholesky GPTQ technology, suitable for text, MTP and visual tasks.

• Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6-i1-GGUF↗

  mradermacher

  This model is a weighted/imatrix quantized version of Qwen3.5-9B-Brainwaves-Coder-Flash-OneJev1.6, provided by nightmedia.

• tinyllama-hatexplain-lora↗

  venkateshmedasani

  A LoRA adapter (plus a classification head), converting TinyLlama/TinyLlama-1.1B-Chat-v1.0 into a three-class hate speech classifier (hate, normal, offensive), trained on HateXplain.