DeepSeek-R1-Distill-Llama-70B is a 70B-parameter distilled LLM, combining reasoning, speed, and accuracy for code, math, and complex logic tasks.
Gemma 3 is a lightweight, multimodal open model by Google with 128K context, multilingual support, and broad deployment versatility.
GPT-OSS-120B is an open-weight 117B-parameter Mixture-of-Experts model by OpenAI, using only 5.1B active parameters per token. Supports reasoning, chain-of-thought, tool use, and fine-tuning.
LLaMA 3.1 8B is a multilingual model optimized for dialogue, outperforming many open and closed-source chat models.
Meta's multilingual 70B parameter language model, optimized for instruction-based dialogue and benchmarking.
Llama-Guard-3-8B is a fine-tuned Llama 3.1-8B model for multilingual content safety classification, detecting unsafe prompts and responses using MLCommons hazard taxonomy.
Mistral-7b for the Italian language, continued pre-training for Italian on a curated large-scale high-quality corpus. Powered by mii-llm
Mistral-Small-3.2-24B-Instruct-2506: 24B multimodal instruction-tuned model, optimized for reasoning and STEM, supports robust function calling, reduces repetition, handles both text and vision inputs efficiently.
Microsoft's phi-4 combines academic and synthetic datasets to enhance reasoning abilities.
Qwen understands images, videos, and text, reasons deeply, generates structured outputs, localizes objects accurately, acts agentically, fast, reliable, flexible, smart.
Qwen3 is the latest generation of large language models in Qwen series.
A 30 B‑parameter MoE coding model, optimized for agentic coding with long 256 K token context, supports function‑call format and strong instruction‑based reasoning
High-performance speech-to-text model based on OpenAI’s Whisper large-v3 architecture, optimized for fast and efficient inference using CTranslate2. It supports multilingual transcription and translation with high accuracy, especially for long-form audio and noisy environments
Qwen2 GTE Embedding
Qwen3 Embedding: advanced multilingual models for text embedding, ranking, retrieval, classification, and clustering.
An image generation foundation model in the Qwen series that achieves significant advances in complex text rendering and precise image editing
