Best AI Models for Linux

LM-Kit.NET now lets your .NET apps run the latest open models entirely on device, including Meta Llama 4, DeepSeek V3-0324, Microsoft Phi 4 (plus mini and multimodal variants), Mistral Mixtral 8x22B, Google Gemma 3, and Alibaba Qwen 2.5 VL, so you get cutting-edge language, vision, and audio performance without calling any external service. A continuously updated model catalog with setup instructions and quantized builds is available at docs.lm-kit.com/lm-kit-net/guides/getting-started/model-catalog.html, letting you integrate new releases quickly while keeping latency low and data fully private.

DeepSeek is a cutting-edge AI assistant powered by the advanced DeepSeek-V3 model, featuring over 600 billion parameters for exceptional performance. Designed to compete with top global AI systems, it offers fast responses and a wide range of features to make everyday tasks easier and more efficient. Available across multiple platforms, including iOS, Android, and the web, DeepSeek ensures accessibility for users everywhere. The app supports multiple languages and has been continually updated to improve functionality, add new language options, and resolve issues. With its seamless performance and versatility, DeepSeek has garnered positive feedback from users worldwide.

Mistral AI is a pioneering artificial intelligence startup specializing in open-source generative AI. The company offers a range of customizable, enterprise-grade AI solutions deployable across various platforms, including on-premises, cloud, edge, and devices. Flagship products include "Le Chat," a multilingual AI assistant designed to enhance productivity in both personal and professional contexts, and "La Plateforme," a developer platform that enables the creation and deployment of AI-powered applications. Committed to transparency and innovation, Mistral AI positions itself as a leading independent AI lab, contributing significantly to open-source AI and policy development.

Cohere is an enterprise AI platform that enables developers and businesses to build powerful language-based applications. Specializing in large language models (LLMs), Cohere provides solutions for text generation, summarization, and semantic search. Their model offerings include the Command family for high-performance language tasks and Aya Expanse for multilingual applications across 23 languages. Focused on security and customization, Cohere allows flexible deployment across major cloud providers, private cloud environments, or on-premises setups to meet diverse enterprise needs. The company collaborates with industry leaders like Oracle and Salesforce to integrate generative AI into business applications, improving automation and customer engagement. Additionally, Cohere For AI, their research lab, advances machine learning through open-source projects and a global research community.

GPT-4.5 is a powerful AI model that improves upon its predecessor by scaling unsupervised learning, enhancing reasoning abilities, and offering improved collaboration capabilities. Designed to better understand human intent and collaborate in more natural, intuitive ways, GPT-4.5 delivers higher accuracy and lower hallucination rates across a broad range of topics. Its advanced capabilities enable it to generate creative and insightful content, solve complex problems, and assist with tasks in writing, design, and even space exploration. With improved AI-human interactions, GPT-4.5 is optimized for practical applications, making it more accessible and reliable for businesses and developers.

Qwen-7B is the 7B-parameter version of the large language model series, Qwen (abbr. Tongyi Qianwen), proposed by Alibaba Cloud. Qwen-7B is a Transformer-based large language model, which is pretrained on a large volume of data, including web texts, books, codes, etc. Additionally, based on the pretrained Qwen-7B, we release Qwen-7B-Chat, a large-model-based AI assistant, which is trained with alignment techniques. The features of the Qwen-7B series include: Trained with high-quality pretraining data. We have pretrained Qwen-7B on a self-constructed large-scale high-quality dataset of over 2.2 trillion tokens. The dataset includes plain texts and codes, and it covers a wide range of domains, including general domain data and professional domain data. Strong performance. In comparison with the models of the similar model size, we outperform the competitors on a series of benchmark datasets, which evaluates natural language understanding, mathematics, coding, etc. And more.

Mistral 7B is a 7.3-billion-parameter language model that outperforms larger models like Llama 2 13B across various benchmarks. It employs Grouped-Query Attention (GQA) for faster inference and Sliding Window Attention (SWA) to efficiently handle longer sequences. Released under the Apache 2.0 license, Mistral 7B is accessible for deployment across diverse platforms, including local environments and major cloud services. Additionally, a fine-tuned version, Mistral 7B Instruct, demonstrates enhanced performance in instruction-following tasks, surpassing models like Llama 2 13B Chat.

Mistral AI introduced two state-of-the-art models for on-device computing and edge use cases, named "les Ministraux": Ministral 3B and Ministral 8B. These models set a new frontier in knowledge, commonsense reasoning, function-calling, and efficiency in the sub-10B category. They can be used or tuned for various applications, from orchestrating agentic workflows to creating specialist task workers. Both models support up to 128k context length (currently 32k on vLLM), and Ministral 8B features a special interleaved sliding-window attention pattern for faster and memory-efficient inference. These models were built to provide a compute-efficient and low-latency solution for scenarios such as on-device translation, internet-less smart assistants, local analytics, and autonomous robotics. Used in conjunction with larger language models like Mistral Large, les Ministraux also serve as efficient intermediaries for function-calling in multi-step agentic workflows.

Mistral AI has introduced two advanced models for on-device computing and edge applications, named "les Ministraux": Ministral 3B and Ministral 8B. These models excel in knowledge, commonsense reasoning, function-calling, and efficiency within the sub-10B parameter range. They support up to 128k context length and are designed for various applications, including on-device translation, offline smart assistants, local analytics, and autonomous robotics. Ministral 8B features an interleaved sliding-window attention pattern for faster and more memory-efficient inference. Both models can function as intermediaries in multi-step agentic workflows, handling tasks like input parsing, task routing, and API calls based on user intent with low latency and cost. Benchmark evaluations indicate that les Ministraux consistently outperforms comparable models across multiple tasks. As of October 16, 2024, both models are available, with Ministral 8B priced at $0.1 per million tokens.

On September 17, 2024, Mistral AI announced several key updates to enhance the accessibility and performance of their AI offerings. They introduced a free tier on "La Plateforme," their serverless platform for tuning and deploying Mistral models as API endpoints, enabling developers to experiment and prototype at no cost. Additionally, Mistral AI reduced prices across their entire model lineup, with significant cuts such as a 50% reduction for Mistral Nemo and an 80% decrease for Mistral Small and Codestral, making advanced AI more cost-effective for users. The company also unveiled Mistral Small v24.09, a 22-billion-parameter model offering a balance between performance and efficiency, suitable for tasks like translation, summarization, and sentiment analysis. Furthermore, they made Pixtral 12B, a vision-capable model with image understanding capabilities, freely available on "Le Chat," allowing users to analyze and caption images without compromising text-based performance.

Grok Code Fast 1 is a high-speed, economical reasoning model designed specifically for agentic coding workflows. Unlike traditional models that can feel slow in tool-based loops, it delivers near-instant responses, excelling in everyday software development tasks. Built from scratch with a programming-rich corpus and refined on real-world pull requests, it supports languages like TypeScript, Python, Java, Rust, C++, and Go. Developers can use it for everything from zero-to-one project building to precise bug fixes and codebase Q&A. With optimized inference and caching techniques, it achieves impressive responsiveness and a 90%+ cache hit rate when integrated with partners like GitHub Copilot, Cursor, and Cline. Offered at just $0.20 per million input tokens and $1.50 per million output tokens, Grok Code Fast 1 strikes a strong balance between speed, performance, and affordability.

Announcing the launch of Jurassic-2, the latest generation of AI21 Studio’s foundation models, a game-changer in the field of AI, with top-tier quality and new capabilities. And that's not all, we're also releasing our task-specific APIs, with plug-and-play reading and writing capabilities that outperform competitors. Our focus at AI21 Studio is to help developers and businesses leverage reading and writing AI to build real-world products with tangible value. Today marks two important milestones with the release of Jurassic-2 and Task-Specific APIs, empowering you to bring generative AI to production. Jurassic-2 (or J2, as we like to call it) is the next generation of our foundation models with significant improvements in quality and new capabilities including zero-shot instruction-following, reduced latency, and multi-language support. Task-specific APIs provide developers with industry-leading APIs that perform specialized reading and writing tasks out-of-the-box.

10x productivity with customizable AI assistants, global hotkeys, and in-line AI. Seamless integration into your mobile workflows with elegant features. Conversations, preferences, and model usage stay on your computer—secure, exportable, and can be deleted at any time.

FLUX.1 is a groundbreaking suite of open-source text-to-image models developed by Black Forest Labs, setting new benchmarks in AI-generated imagery with its 12 billion parameters. It surpasses established models like Midjourney V6, DALL-E 3, and Stable Diffusion 3 Ultra by offering superior image quality, detail, prompt fidelity, and versatility across various styles and scenes. FLUX.1 comes in three variants: Pro for top-tier commercial use, Dev for non-commercial research with efficiency akin to Pro, and Schnell for rapid personal and local development projects under an Apache 2.0 license. Its innovative use of flow matching and rotary positional embeddings allows for efficient and high-quality image synthesis, making FLUX.1 a significant advancement in the domain of AI-driven visual creativity.

HunyuanVideo‑Avatar supports animating any input avatar images to high‑dynamic, emotion‑controllable videos using simple audio conditions. It is a multimodal diffusion transformer (MM‑DiT)‑based model capable of generating dynamic, emotion‑controllable, multi‑character dialogue videos. It accepts multi‑style avatar inputs, photorealistic, cartoon, 3D‑rendered, anthropomorphic, at arbitrary scales from portrait to full body. Provides a character image injection module that ensures strong character consistency while enabling dynamic motion; an Audio Emotion Module (AEM) that extracts emotional cues from a reference image to enable fine‑grained emotion control over generated video; and a Face‑Aware Audio Adapter (FAA) that isolates audio influence to specific face regions via latent‑level masking, supporting independent audio‑driven animation in multi‑character scenarios.

Qwen3‑Coder is an agentic code model available in multiple sizes, led by the 480B‑parameter Mixture‑of‑Experts variant (35B active) that natively supports 256K‑token contexts (extendable to 1M) and achieves state‑of‑the‑art results comparable to Claude Sonnet 4. Pre‑training on 7.5T tokens (70 % code) and synthetic data cleaned via Qwen2.5‑Coder optimized both coding proficiency and general abilities, while post‑training employs large‑scale, execution‑driven reinforcement learning, scaling test‑case generation for diverse coding challenges, and long‑horizon RL across 20,000 parallel environments to excel on multi‑turn software‑engineering benchmarks like SWE‑Bench Verified without test‑time scaling. Alongside the model, the open source Qwen Code CLI (forked from Gemini Code) unleashes Qwen3‑Coder in agentic workflows with customized prompts, function calling protocols, and seamless integration with Node.js, OpenAI SDKs, and environment variables.

HunyuanVision is a cutting-edge vision-language model developed by Tencent’s Hunyuan team. It uses a mamba-transformer hybrid architecture to deliver strong performance and efficient inference in multimodal reasoning tasks. The version Hunyuan-Vision-1.5 is designed for “thinking on images,” meaning it not only understands vision+language content, but can perform deeper reasoning that involves manipulating or reflecting on image inputs, such as cropping, zooming, pointing, box drawing, or drawing on the image to acquire additional knowledge. It supports a variety of vision tasks (image + video recognition, OCR, diagram understanding), visual reasoning, and even 3D spatial comprehension, all in a unified multilingual framework. The model is built to work seamlessly across languages and tasks and is intended to be open sourced (including checkpoints, technical report, inference support) to encourage the community to experiment and adopt.

Yi-Large is a proprietary large language model developed by 01.AI, offering a 32k context length with both input and output costs at $2 per million tokens. It stands out with its advanced capabilities in natural language processing, common-sense reasoning, and multilingual support, performing on par with leading models like GPT-4 and Claude3 in various benchmarks. Yi-Large is designed for tasks requiring complex inference, prediction, and language understanding, making it suitable for applications like knowledge search, data classification, and creating human-like chatbots. Its architecture is based on a decoder-only transformer with enhancements such as pre-normalization and Group Query Attention, and it has been trained on a vast, high-quality multilingual dataset. This model's versatility and cost-efficiency make it a strong contender in the AI market, particularly for enterprises aiming to deploy AI solutions globally.

PaLM API is an easy and safe way to build on top of our best language models. Today, we’re making an efficient model available, in terms of size and capabilities, and we’ll add other sizes soon. The API also comes with an intuitive tool called MakerSuite, which lets you quickly prototype ideas and, over time, will have features for prompt engineering, synthetic data generation and custom-model tuning — all supported by robust safety tools. Select developers can access the PaLM API and MakerSuite in Private Preview today, and stay tuned for our waitlist soon.

gpt-oss-20b is a 20-billion-parameter, text-only reasoning model released under the Apache 2.0 license and governed by OpenAI’s gpt-oss usage policy, built to enable seamless integration into custom AI workflows via the Responses API without reliance on proprietary infrastructure. Trained for robust instruction following, it supports adjustable reasoning effort, full chain-of-thought outputs, and native tool use (including web search and Python execution), producing structured, explainable answers. Developers must implement their own deployment safeguards, such as input filtering, output monitoring, and usage policies, to match the system-level protections of hosted offerings and mitigate risks from malicious or unintended behaviors. Its open-weight design makes it ideal for on-premises or edge deployments where control, customization, and transparency are paramount.

gpt-oss-120b is a reasoning model engineered for deep, transparent thinking, delivering full chain-of-thought explanations, adjustable reasoning depth, and structured outputs, while natively invoking tools like web search and Python execution via the API. Built to slot seamlessly into self-hosted or edge deployments, it eliminates dependence on proprietary infrastructure. Although it includes default safety guardrails, its open-weight architecture allows fine-tuning that could override built-in controls, so implementers are responsible for adding input filtering, output monitoring, and governance measures to achieve enterprise-grade security. As a community–driven model card rather than a managed service spec, it emphasizes transparency, customization, and the need for downstream safety practices.