Discover AI models for every task
Showing 1-16 of 16 models
by Google
Gemma 3 27B IT is a cutting-edge multimodal vision-language model with 27 billion parameters, built on Gemini technology. Trained on 14 trillion tokens, it handles both text and image inputs with a 128K context window and supports 140+ languages. Excels at visual understanding, code generation, mathematical reasoning, and multilingual tasks. Achieves 78.6 on MMLU, 82.6 on GSM8K, 85.6 on DocVQA, and 76.3 on ChartQA. Lightweight enough for laptop deployment with strong safety improvements over previous Gemma versions.
Larger Gemma model delivering high-quality chat and coding with efficient inference.
Google Gemma 4 31B is a 31B parameter dense multimodal language model with a 256K context window. It processes text, images, and video inputs and generates text output, featuring a configurable thinking mode for step‑by‑step reasoning. The model achieves 85.2% on MMLU Pro, 80.0% on LiveCodeBench v6, and 88.4% on MMMLU, demonstrating strong performance across reasoning and multimodal benchmarks. Available under the Apache 2.0 license.
by BAAI
BAAI BGE Multilingual Gemma2 is a multilingual dense retrieval embedding model built on Gemma 2 architecture, supporting 100+ languages for cross-lingual semantic search and retrieval. Delivers strong performance across diverse language families including English, Chinese, Spanish, Arabic, Hindi, and many more. Ideal for multilingual search systems, cross-lingual document retrieval, international content recommendation, and global knowledge bases. Trained on large-scale multilingual data with balanced language representation.
by Mistral
Mistral Small 3.2 24B Instruct is a multimodal instruction-tuned model supporting both vision and text with 24B parameters and 128K context. Major improvements over 3.1 include better instruction following (84.78%), 2x reduction in repetition errors, and robust function calling. Achieves 65.33% on Wildbench v2, 43.1% on Arena Hard v2, 92.90% on HumanEval Pass@5. Vision benchmarks: 87.4% ChartQA, 94.86% DocVQA, 62.50% MMMU. Supports up to 10 images per prompt with integrated vision-based function calling.
by Meta
Meta Llama 3.3 70B Instruct is a multilingual instruction-tuned model optimized for dialogue. Trained on ~15 trillion tokens with cutoff December 2023, it outperforms many open-source and closed models. Major improvements include 92.1% on IFEval (steerability), 88.4% on HumanEval (code), 77.0% on MATH, and 91.1% on MGSM (multilingual). Features 128K context, Grouped-Query Attention, and supports 8 languages including English, German, French, Spanish, Italian, Portuguese, Hindi, and Thai. Trained on 7M GPU hours with 100% renewable energy.
Moderation model providing robust safety classification and policy enforcement.
Meta Llama 3.1 8B Instruct is an efficient multilingual instruction-tuned model optimized for dialogue and assistant use cases. With 8 billion parameters and 128K context length, it provides strong performance across general tasks, code generation, and multilingual understanding. Supports function calling and tool use with Grouped-Query Attention architecture. Ideal for deployment scenarios requiring lower compute resources while maintaining quality across English and 7 additional languages including German, French, Spanish, and Hindi.
Meta's flagship 405B parameter model representing the pinnacle of open-source AI. Exceptional reasoning and comprehensive knowledge for demanding applications.
by NVIDIA
NVIDIA Nemotron 3 Super 120B A12B FP8 is a 120B parameter (12B active) LatentMixture-of-Experts hybrid model with Mamba-2, MoE and Multi-Token Prediction layers, supporting up to 1M tokens context. It achieves 94.73% on HMMT Feb25 (with tools) and 83.73% on MMLU‑Pro, and scores 73.88% on Arena‑Hard‑V2 (Hard Prompt). The model supports configurable reasoning via an enable_thinking flag, tool use, and structured output. It is available under the NVIDIA Nemotron Open Model License.
by intfloat
intfloat E5-Mistral-7B-Instruct is a state-of-the-art instruction-following embedding model built on Mistral 7B architecture. Combines strong language understanding from Mistral with specialized embedding training for retrieval tasks. Features instruction-based embedding generation allowing natural language queries to guide semantic search. Excels at complex retrieval scenarios, multi-hop reasoning in document search, and instruction-guided similarity tasks. Provides significantly improved zero-shot retrieval performance compared to traditional embedding models.
by PrimeIntellect
INTELLECT-3 is an open-source foundation model released under Apache 2.0 license. Designed for general-purpose tasks including text generation, reasoning, and code understanding. Built through decentralized training approach, representing community-driven AI development. Features strong performance across diverse benchmarks while maintaining accessibility for research and commercial use. Supports function calling and tool integration with focus on practical deployment scenarios.
by Qwen
Qwen 3.5 9B is a 9B‑parameter multimodal large language model with a gated‑delta mixture‑of‑experts architecture and a vision encoder. It supports a native context window of 262,144 tokens and operates in a default thinking mode that can be disabled. The model achieves strong results such as 82.5% on MMLU‑Pro, 88.2% on C‑Eval, and 78.4% on MMMU benchmarks. It is released under the Apache 2.0 license.
NVIDIA Nemotron 3 Nano is a highly efficient hybrid Mamba-Transformer MoE model with 30B total / 3.5B active parameters. Features 128K context window extensible to 1M tokens. Excels at agentic AI, reasoning, and tool calling tasks. Trained on 25T tokens with state-of-the-art efficiency. Supports English, German, French, Spanish, Italian, and Japanese. Open weights with commercial license.
Qwen 3.5 397B A17B is a 397B-parameter mixture-of-experts vision-language foundation model with a gated delta network architecture and a vision encoder. It supports a native context window of 262,144 tokens (extendable to over 1 million) and operates in a default thinking mode that can be disabled. The model achieves strong results such as 87.8% on MMLU‑Pro, 85.0% on MMMU, and 88.6% on MathVision benchmarks. It is released under the Apache 2.0 license.
Specialized 13B coding model with advanced code infilling capabilities. Excels in code generation, completion, debugging across multiple programming languages.