Top Ministral 8B Alternatives

Mistral Small 3.1 is an advanced, multimodal, and multilingual AI model that has been made available under the Apache 2.0 license. Building upon the previous Mistral Small 3, this updated version showcases improved text processing abilities and enhanced multimodal understanding, with the capacity to handle an extensive context window of up to 128,000 tokens. It outperforms comparable models like Gemma 3 and GPT-4o Mini, reaching remarkable inference rates of 150 tokens per second. Designed for versatility, Mistral Small 3.1 excels in various applications, including instruction adherence, conversational interaction, visual data interpretation, and executing functions, making it suitable for both commercial and individual AI uses. Its efficient architecture allows it to run smoothly on hardware configurations such as a single RTX 4090 or a Mac with 32GB of RAM, enabling on-device operations. Users have the option to download the model from Hugging Face and explore its features via Mistral AI's developer playground, while it is also embedded in services like Google Cloud Vertex AI and accessible on platforms like NVIDIA NIM. This extensive flexibility empowers developers to utilize its advanced capabilities across a wide range of environments and applications, thereby maximizing its potential impact in the AI landscape. Furthermore, Mistral Small 3.1's innovative design ensures that it remains adaptable to future technological advancements.

Mistral AI has introduced two state-of-the-art models aimed at on-device computing and edge applications, collectively known as "les Ministraux": Ministral 3B and Ministral 8B. These advanced models set new benchmarks for knowledge, commonsense reasoning, function-calling, and efficiency in the sub-10B category. They offer remarkable flexibility for a variety of applications, from overseeing complex workflows to creating specialized task-oriented agents. With the capability to manage an impressive context length of up to 128k (currently supporting 32k on vLLM), Ministral 8B features a distinctive interleaved sliding-window attention mechanism that boosts both speed and memory efficiency during inference. Crafted for low-latency and compute-efficient applications, these models thrive in environments such as offline translation, internet-independent smart assistants, local data processing, and autonomous robotics. Additionally, when integrated with larger language models like Mistral Large, les Ministraux can serve as effective intermediaries, enhancing function-calling within detailed multi-step workflows. This synergy not only amplifies performance but also extends the potential of AI in edge computing, paving the way for innovative solutions in various fields. The introduction of these models marks a significant step forward in making advanced AI more accessible and efficient for real-world applications.

Mistral Large is the flagship language model developed by Mistral AI, designed for advanced text generation and complex multilingual reasoning tasks including text understanding, transformation, and software code creation. It supports various languages such as English, French, Spanish, German, and Italian, enabling it to effectively navigate grammatical complexities and cultural subtleties. With a remarkable context window of 32,000 tokens, Mistral Large can accurately retain and reference information from extensive documents. Its proficiency in following precise instructions and invoking built-in functions significantly aids in application development and the modernization of technology infrastructures. Accessible through Mistral's platform, Azure AI Studio, and Azure Machine Learning, it also provides an option for self-deployment, making it suitable for sensitive applications. Benchmark results indicate that Mistral Large excels in performance, ranking as the second-best model worldwide available through an API, closely following GPT-4, which underscores its strong position within the AI sector. This blend of features and capabilities positions Mistral Large as an essential resource for developers aiming to harness cutting-edge AI technologies effectively. Moreover, its adaptable nature allows it to meet diverse industry needs, further enhancing its appeal as a versatile AI solution.

LFM2 is a cutting-edge series of on-device foundation models specifically engineered to deliver an exceptionally fast generative-AI experience across a wide range of devices. It employs an innovative hybrid architecture that enables decoding and pre-filling speeds up to twice as fast as competing models, while also improving training efficiency by as much as threefold compared to earlier versions. Striking a perfect balance between quality, latency, and memory use, these models are ideally suited for embedded system applications, allowing for real-time, on-device AI capabilities in smartphones, laptops, vehicles, wearables, and many other platforms. This results in millisecond-level inference, enhanced device longevity, and complete data sovereignty for users. Available in three configurations with 0.35 billion, 0.7 billion, and 1.2 billion parameters, LFM2 demonstrates superior benchmark results compared to similarly sized models, excelling in knowledge recall, mathematical problem-solving, adherence to multilingual instructions, and conversational dialogue evaluations. With such impressive capabilities, LFM2 not only elevates the user experience but also establishes a new benchmark for on-device AI performance, paving the way for future advancements in the field.

Ai2 OLMoE is a completely open-source language model that utilizes a mixture-of-experts approach, designed to operate fully on-device, which allows users to explore its capabilities in a secure and private environment. The primary goal of this application is to aid researchers in enhancing on-device intelligence while enabling developers to rapidly prototype innovative AI applications without relying on cloud services. As a highly efficient version within the Ai2 OLMo model family, OLMoE empowers users to engage with advanced local models in practical situations, explore strategies to improve smaller AI systems, and locally test their models using the provided open-source framework. Furthermore, OLMoE can be smoothly integrated into a variety of iOS applications, prioritizing user privacy and security by functioning entirely on-device. Users can easily share the results of their conversations with friends or colleagues, enjoying the benefits of a completely open-source model and application code. This makes Ai2 OLMoE an outstanding resource for personal experimentation and collaborative research, offering extensive opportunities for innovation and discovery in the field of artificial intelligence. By leveraging OLMoE, users can contribute to a growing ecosystem of on-device AI solutions that respect user privacy while facilitating cutting-edge advancements.

Mistral 7B is a cutting-edge language model boasting 7.3 billion parameters, which excels in various benchmarks, even surpassing larger models such as Llama 2 13B. It employs advanced methods like Grouped-Query Attention (GQA) to enhance inference speed and Sliding Window Attention (SWA) to effectively handle extensive sequences. Available under the Apache 2.0 license, Mistral 7B can be deployed across multiple platforms, including local infrastructures and major cloud services. Additionally, a unique variant called Mistral 7B Instruct has demonstrated exceptional abilities in task execution, consistently outperforming rivals like Llama 2 13B Chat in certain applications. This adaptability and performance make Mistral 7B a compelling choice for both developers and researchers seeking efficient solutions. Its innovative features and strong results highlight the model's potential impact on natural language processing projects.

Mistral AI has unveiled the Mistral Large 2, an advanced AI model engineered to perform exceptionally well across various fields, including code generation, multilingual comprehension, and complex reasoning tasks. Boasting a remarkable 128k context window, this model supports a vast selection of languages such as English, French, Spanish, and Arabic, as well as more than 80 programming languages. Tailored for high-throughput single-node inference, Mistral Large 2 is ideal for applications that demand substantial context management. Its outstanding performance on benchmarks like MMLU, alongside enhanced abilities in code generation and reasoning, ensures both precision and effectiveness in outcomes. Moreover, the model is equipped with improved function calling and retrieval functionalities, which are especially advantageous for intricate business applications. This versatility positions Mistral Large 2 as a formidable asset for developers and enterprises eager to harness cutting-edge AI technologies for innovative solutions, ultimately driving efficiency and productivity in their operations.

Z.ai has launched its newest flagship model, GLM-4.5, which features an astounding total of 355 billion parameters (with 32 billion actively utilized) and is accompanied by the GLM-4.5-Air variant, which includes 106 billion parameters (12 billion active) tailored for advanced reasoning, coding, and agent-like functionalities within a unified framework. This innovative model is capable of toggling between a "thinking" mode, ideal for complex, multi-step reasoning and tool utilization, and a "non-thinking" mode that allows for quick responses, supporting a context length of up to 128K tokens and enabling native function calls. Available via the Z.ai chat platform and API, and with open weights on sites like HuggingFace and ModelScope, GLM-4.5 excels at handling diverse inputs for various tasks, including general problem solving, common-sense reasoning, coding from scratch or enhancing existing frameworks, and orchestrating extensive workflows such as web browsing and slide creation. The underlying architecture employs a Mixture-of-Experts design that incorporates loss-free balance routing, grouped-query attention mechanisms, and an MTP layer to support speculative decoding, ensuring it meets enterprise-level performance expectations while being versatile enough for a wide array of applications. Consequently, GLM-4.5 sets a remarkable standard for AI capabilities, pushing the boundaries of technology across multiple fields and industries. This advancement not only enhances user experience but also drives innovation in artificial intelligence solutions.

Yi-Large is a cutting-edge proprietary large language model developed by 01.AI, boasting an impressive context length of 32,000 tokens and a pricing model set at $2 per million tokens for both input and output. Celebrated for its exceptional capabilities in natural language processing, common-sense reasoning, and multilingual support, it stands out in competition with leading models like GPT-4 and Claude3 in diverse assessments. The model excels in complex tasks that demand deep inference, precise prediction, and thorough language understanding, making it particularly suitable for applications such as knowledge retrieval, data classification, and the creation of conversational chatbots that closely resemble human communication. Utilizing a decoder-only transformer architecture, Yi-Large integrates advanced features such as pre-normalization and Group Query Attention, having been trained on a vast, high-quality multilingual dataset to optimize its effectiveness. Its versatility and cost-effective pricing make it a powerful contender in the realm of artificial intelligence, particularly for organizations aiming to adopt AI technologies on a worldwide scale. Furthermore, its adaptability across various applications highlights its potential to transform how businesses utilize language models for an array of requirements, paving the way for innovative solutions in the industry. Thus, Yi-Large not only meets but also exceeds expectations, solidifying its role as a pivotal tool in the advancements of AI-driven communication.

We are excited to unveil Mistral NeMo, our latest and most sophisticated small model, boasting an impressive 12 billion parameters and a vast context length of 128,000 tokens, all available under the Apache 2.0 license. In collaboration with NVIDIA, Mistral NeMo stands out in its category for its exceptional reasoning capabilities, extensive world knowledge, and coding skills. Its architecture adheres to established industry standards, ensuring it is user-friendly and serves as a smooth transition for those currently using Mistral 7B. To encourage adoption by researchers and businesses alike, we are providing both pre-trained base models and instruction-tuned checkpoints, all under the Apache license. A remarkable feature of Mistral NeMo is its quantization awareness, which enables FP8 inference while maintaining high performance levels. Additionally, the model is well-suited for a range of global applications, showcasing its ability in function calling and offering a significant context window. When benchmarked against Mistral 7B, Mistral NeMo demonstrates a marked improvement in comprehending and executing intricate instructions, highlighting its advanced reasoning abilities and capacity to handle complex multi-turn dialogues. Furthermore, its design not only enhances its performance but also positions it as a formidable option for multi-lingual tasks, ensuring it meets the diverse needs of various use cases while paving the way for future innovations.

On September 17, 2024, Mistral AI announced a series of important enhancements aimed at making their AI products more accessible and efficient. Among these advancements, they introduced a free tier on "La Plateforme," their serverless platform that facilitates the tuning and deployment of Mistral models as API endpoints, enabling developers to experiment and create without any cost. Additionally, Mistral AI implemented significant price reductions across their entire model lineup, featuring a striking 50% reduction for Mistral Nemo and an astounding 80% decrease for Mistral Small and Codestral, making sophisticated AI solutions much more affordable for a larger audience. Furthermore, the company unveiled Mistral Small v24.09, a model boasting 22 billion parameters, which offers an excellent balance between performance and efficiency, suitable for a range of applications such as translation, summarization, and sentiment analysis. They also launched Pixtral 12B, a vision-capable model with advanced image understanding functionalities, available for free on "Le Chat," which allows users to analyze and caption images while ensuring strong text-based performance. These updates not only showcase Mistral AI's dedication to enhancing their offerings but also underscore their mission to make cutting-edge AI technology accessible to developers across the globe. This commitment to accessibility and innovation positions Mistral AI as a leader in the AI industry.

Reka Flash 3 stands as a state-of-the-art multimodal AI model, boasting 21 billion parameters and developed by Reka AI, to excel in diverse tasks such as engaging in general conversations, coding, adhering to instructions, and executing various functions. This innovative model skillfully processes and interprets a wide range of inputs, which includes text, images, video, and audio, making it a compact yet versatile solution fit for numerous applications. Constructed from the ground up, Reka Flash 3 was trained on a diverse collection of datasets that include both publicly accessible and synthetic data, undergoing a thorough instruction tuning process with carefully selected high-quality information to refine its performance. The concluding stage of its training leveraged reinforcement learning techniques, specifically the REINFORCE Leave One-Out (RLOO) method, which integrated both model-driven and rule-oriented rewards to enhance its reasoning capabilities significantly. With a remarkable context length of 32,000 tokens, Reka Flash 3 effectively competes against proprietary models such as OpenAI's o1-mini, making it highly suitable for applications that demand low latency or on-device processing. Operating at full precision, the model requires a memory footprint of 39GB (fp16), but this can be optimized down to just 11GB through 4-bit quantization, showcasing its flexibility across various deployment environments. Furthermore, Reka Flash 3's advanced features ensure that it can adapt to a wide array of user requirements, thereby reinforcing its position as a leader in the realm of multimodal AI technology. This advancement not only highlights the progress made in AI but also opens doors to new possibilities for innovation across different sectors.

Llama 4 Scout represents a leap forward in multimodal AI, featuring 17 billion active parameters and a groundbreaking 10 million token context length. With its ability to integrate both text and image data, Llama 4 Scout excels at tasks like multi-document summarization, complex reasoning, and image grounding. It delivers superior performance across various benchmarks and is particularly effective in applications requiring both language and visual comprehension. Scout's efficiency and advanced capabilities make it an ideal solution for developers and businesses looking for a versatile and powerful model to enhance their AI-driven projects.

The Megatron-Turing Natural Language Generation model (MT-NLG) is distinguished as the most extensive and sophisticated monolithic transformer model designed for the English language, featuring an astounding 530 billion parameters. Its architecture, consisting of 105 layers, significantly amplifies the performance of prior top models, especially in scenarios involving zero-shot, one-shot, and few-shot learning. The model demonstrates remarkable accuracy across a diverse array of natural language processing tasks, such as completion prediction, reading comprehension, commonsense reasoning, natural language inference, and word sense disambiguation. In a bid to encourage further exploration of this revolutionary English language model and to enable users to harness its capabilities across various linguistic applications, NVIDIA has launched an Early Access program that offers a managed API service specifically for the MT-NLG model. This program is designed not only to promote experimentation but also to inspire innovation within the natural language processing domain, ultimately paving the way for new advancements in the field. Through this initiative, researchers and developers will have the opportunity to delve deeper into the potential of MT-NLG and contribute to its evolution.

Fullmoon stands out as a groundbreaking, open-source app that empowers users to interact directly with large language models right on their personal devices, emphasizing user privacy and offline capabilities. Specifically optimized for Apple silicon, it operates efficiently across a range of platforms, including iOS, iPadOS, macOS, and visionOS, ensuring a cohesive user experience. Users can tailor their interactions by adjusting themes, fonts, and system prompts, and the app’s integration with Apple’s Shortcuts further boosts productivity. Importantly, Fullmoon supports models like Llama-3.2-1B-Instruct-4bit and Llama-3.2-3B-Instruct-4bit, facilitating robust AI engagements without the need for an internet connection. This unique combination of features positions Fullmoon as a highly adaptable tool for individuals seeking to leverage AI technology conveniently and securely. Additionally, the app's emphasis on customization allows users to create an environment that perfectly suits their preferences and needs.

The newest version of the open-source AI framework, which can be customized and utilized across different platforms, is available in several configurations: 1B, 3B, 11B, and 90B, while still offering the option to use Llama 3.1. Llama 3.2 includes a selection of large language models (LLMs) that are pretrained and fine-tuned specifically for multilingual text processing in 1B and 3B sizes, whereas the 11B and 90B models support both text and image inputs, generating text outputs. This latest release empowers users to build highly effective applications that cater to specific requirements. For applications running directly on devices, such as summarizing conversations or managing calendars, the 1B or 3B models are excellent selections. On the other hand, the 11B and 90B models are particularly suited for tasks involving images, allowing users to manipulate existing pictures or glean further insights from images in their surroundings. Ultimately, this broad spectrum of models opens the door for developers to experiment with creative applications across a wide array of fields, enhancing the potential for innovation and impact.

DeepSeek-V2 represents an advanced Mixture-of-Experts (MoE) language model created by DeepSeek-AI, recognized for its economical training and superior inference efficiency. This model features a staggering 236 billion parameters, engaging only 21 billion for each token, and can manage a context length stretching up to 128K tokens. It employs sophisticated architectures like Multi-head Latent Attention (MLA) to enhance inference by reducing the Key-Value (KV) cache and utilizes DeepSeekMoE for cost-effective training through sparse computations. When compared to its earlier version, DeepSeek 67B, this model exhibits substantial advancements, boasting a 42.5% decrease in training costs, a 93.3% reduction in KV cache size, and a remarkable 5.76-fold increase in generation speed. With training based on an extensive dataset of 8.1 trillion tokens, DeepSeek-V2 showcases outstanding proficiency in language understanding, programming, and reasoning tasks, thereby establishing itself as a premier open-source model in the current landscape. Its groundbreaking methodology not only enhances performance but also sets unprecedented standards in the realm of artificial intelligence, inspiring future innovations in the field.

Mistral Medium 3 is a breakthrough in AI technology, offering the perfect balance of cutting-edge performance and significantly reduced costs. This model introduces a new era of enterprise AI, with a focus on simplifying deployments while still providing exceptional performance. Its ability to deliver high-level results at just a fraction of the cost of its competitors makes it a game-changer in industries that rely on complex AI tasks. Mistral Medium 3 is particularly strong in professional use cases like coding, where it competes closely with larger models that are typically more expensive and slower. The model supports hybrid and on-premises deployments, offering enterprise users full control over customization and integration into their systems. Businesses can leverage Mistral Medium 3 for both large-scale deployments and fine-tuned, domain-specific training, allowing for enhanced efficiency in industries such as healthcare, financial services, and energy. The addition of continuous learning and the ability to integrate with enterprise knowledge bases makes it a flexible, future-proof solution. Customers in beta are already using Mistral Medium 3 to enrich customer service, personalize business processes, and analyze complex datasets, demonstrating its real-world value. Available through various cloud platforms like Amazon Sagemaker, IBM WatsonX, and Google Cloud Vertex, Mistral Medium 3 is now ready to be deployed for custom use cases across a range of industries.

Amazon Nova Premier represents the pinnacle of AI-powered performance, offering capabilities that are essential for high-level tasks that require precise execution, like data synthesis, multi-agent collaboration, and long-form document processing. The model is part of Amazon's Bedrock platform, which integrates with Amazon's ecosystem for seamless AI management. Nova Premier’s one-million token context allows it to process vast amounts of data, making it a powerful tool for handling complex documents, lengthy codebases, and multi-step tasks. It excels at generating accurate, detailed responses, which are crucial in industries like finance and technology, where precision and depth are paramount. As the most advanced model in the Nova family, it can also distill smaller, faster versions of itself, such as Nova Pro and Nova Micro, creating customized models that balance performance with cost-effectiveness for specific use cases. In a real-world application, Nova Premier has been used to enhance investment research workflows, streamlining the data collection process and providing actionable insights faster than ever. This powerful AI tool allows businesses to automate complex processes, enhancing productivity and boosting success rates in critical tasks like proposal writing or data analysis. By leveraging Nova Premier’s capabilities, companies can significantly improve operational efficiency and decision-making accuracy.

Qwen3-Max is Alibaba's state-of-the-art large language model, boasting an impressive trillion parameters designed to enhance performance in tasks that demand agency, coding, reasoning, and the management of long contexts. As a progression of the Qwen3 series, this model utilizes improved architecture, training techniques, and inference methods; it features both thinker and non-thinker modes, introduces a distinctive “thinking budget” approach, and offers the flexibility to switch modes according to the complexity of the tasks. With its capability to process extremely long inputs and manage hundreds of thousands of tokens, it also enables the invocation of tools and showcases remarkable outcomes across various benchmarks, including evaluations related to coding, multi-step reasoning, and agent assessments like Tau2-Bench. Although the initial iteration primarily focuses on following instructions within a non-thinking framework, Alibaba plans to roll out reasoning features that will empower autonomous agent functionalities in the near future. Furthermore, with its robust multilingual support and comprehensive training on trillions of tokens, Qwen3-Max is available through API interfaces that integrate well with OpenAI-style functionalities, guaranteeing extensive applicability across a range of applications. This extensive and innovative framework positions Qwen3-Max as a significant competitor in the field of advanced artificial intelligence language models, making it a pivotal tool for developers and researchers alike.

Kimi K2 Thinking is an advanced open-source reasoning model developed by Moonshot AI, specifically designed for complex, multi-step workflows where it adeptly merges chain-of-thought reasoning with the use of tools across various sequential tasks. It utilizes a state-of-the-art mixture-of-experts architecture, encompassing an impressive total of 1 trillion parameters, though only approximately 32 billion parameters are engaged during each inference, which boosts efficiency while retaining substantial capability. The model supports a context window of up to 256,000 tokens, enabling it to handle extraordinarily lengthy inputs and reasoning sequences without losing coherence. Furthermore, it incorporates native INT4 quantization, which dramatically reduces inference latency and memory usage while maintaining high performance. Tailored for agentic workflows, Kimi K2 Thinking can autonomously trigger external tools, managing sequential logic steps that typically involve around 200-300 tool calls in a single chain while ensuring consistent reasoning throughout the entire process. Its strong architecture positions it as an optimal solution for intricate reasoning challenges that demand both depth and efficiency, making it a valuable asset in various applications. Overall, Kimi K2 Thinking stands out for its ability to integrate complex reasoning and tool use seamlessly.

The ERNIE X1 Turbo by Baidu is a powerful AI model that excels in complex tasks like logical reasoning, text generation, and creative problem-solving. It is designed to process multimodal data, including text and images, making it ideal for a wide range of applications. What sets ERNIE X1 Turbo apart from its competitors is its remarkable performance at an accessible price—just 25% of the cost of the leading models in the market. With its real-time data-driven insights, ERNIE X1 Turbo is perfect for developers, enterprises, and researchers looking to incorporate advanced AI solutions into their workflows without high financial barriers.

A streamlined model that excels in both text comprehension and multimodal reasoning abilities. The GPT-4o mini has been crafted to efficiently manage a vast range of tasks, characterized by its affordability and quick response times, which make it particularly suitable for scenarios requiring the simultaneous execution of multiple model calls, such as activating various APIs at once, analyzing large sets of information like complete codebases or lengthy conversation histories, and delivering prompt, real-time text interactions for customer support chatbots. At present, the API for GPT-4o mini supports both textual and visual inputs, with future enhancements planned to incorporate support for text, images, videos, and audio. This model features an impressive context window of 128K tokens and can produce outputs of up to 16K tokens per request, all while maintaining a knowledge base that is updated to October 2023. Furthermore, the advanced tokenizer utilized in GPT-4o enhances its efficiency in handling non-English text, thus expanding its applicability across a wider range of uses. Consequently, the GPT-4o mini is recognized as an adaptable resource for developers and enterprises, making it a valuable asset in various technological endeavors. Its flexibility and efficiency position it as a leader in the evolving landscape of AI-driven solutions.

Explore the boundaries of personal alignment, artificial intelligence, open-source initiatives, and decentralization through bold experimentation that many large corporations and governmental bodies tend to avoid. Hermes 3 is equipped with advanced features such as robust long-term context retention and the capability to facilitate multi-turn dialogues, alongside complex role-playing and internal monologue functionalities, as well as enhanced agentic function-calling abilities. This model is meticulously designed to ensure accurate compliance with system prompts and instructions while remaining adaptable. By refining Llama 3.1 in various configurations—ranging from 8B to 70B and even 405B—and leveraging a dataset primarily made up of synthetically created examples, Hermes 3 not only matches but often outperforms Llama 3.1, revealing deeper potential for reasoning and innovative tasks. This series of models focused on instruction and tool usage showcases remarkable reasoning and creative capabilities, setting the stage for groundbreaking applications. Ultimately, Hermes 3 signifies a transformative leap in the realm of AI technology, promising to reshape future interactions and developments. As we continue to innovate, the possibilities for practical applications seem boundless.

The Mixtral 8x7B model represents a cutting-edge sparse mixture of experts (SMoE) architecture that features open weights and is made available under the Apache 2.0 license. This innovative model outperforms Llama 2 70B across a range of benchmarks, while also achieving inference speeds that are sixfold faster. As the premier open-weight model with a versatile licensing structure, Mixtral stands out for its impressive cost-effectiveness and performance metrics. Furthermore, it competes with and frequently exceeds the capabilities of GPT-3.5 in many established benchmarks, underscoring its importance in the AI landscape. Its unique blend of accessibility, rapid processing, and overall effectiveness positions it as an attractive option for developers in search of top-tier AI solutions. Consequently, the Mixtral model not only enhances the current technological landscape but also paves the way for future advancements in AI development.

This year marks the 2311th anniversary of Archimedes, and in his honor, we are thrilled to unveil our first Mathstral model, a dedicated 7B architecture crafted specifically for mathematical reasoning and scientific inquiry. With a context window of 32k, this model is made available under the Apache 2.0 license. Our goal in sharing Mathstral with the scientific community is to facilitate the tackling of complex mathematical problems that require sophisticated, multi-step logical reasoning. The introduction of Mathstral aligns with our broader initiative to bolster academic efforts, developed alongside Project Numina. Much like Isaac Newton's contributions during his lifetime, Mathstral builds upon the groundwork established by Mistral 7B, with a keen focus on STEM fields. It showcases exceptional reasoning abilities within its domain, achieving impressive results across numerous industry-standard benchmarks. Specifically, it registers a score of 56.6% on the MATH benchmark and 63.47% on the MMLU benchmark, highlighting the performance enhancements in comparison to its predecessor, Mistral 7B, and underscoring the strides made in mathematical modeling. In addition to advancing individual research, this initiative seeks to inspire greater innovation and foster collaboration within the mathematical community as a whole.

We are excited to present DeepSeek-V3.2-Exp, our latest experimental model that evolves from V3.1-Terminus, incorporating the cutting-edge DeepSeek Sparse Attention (DSA) technology designed to significantly improve both training and inference speeds for longer contexts. This innovative DSA framework enables accurate sparse attention while preserving the quality of outputs, resulting in enhanced performance for long-context tasks alongside reduced computational costs. Benchmark evaluations demonstrate that V3.2-Exp delivers performance on par with V3.1-Terminus, all while benefiting from these efficiency gains. The model is fully functional across various platforms, including app, web, and API. In addition, to promote wider accessibility, we have reduced DeepSeek API pricing by more than 50% starting now. During this transition phase, users will have access to V3.1-Terminus through a temporary API endpoint until October 15, 2025. DeepSeek invites feedback on DSA from users via our dedicated feedback portal, encouraging community engagement. To further support this initiative, DeepSeek-V3.2-Exp is now available as open-source, with model weights and key technologies—including essential GPU kernels in TileLang and CUDA—published on Hugging Face, and we are eager to observe how the community will leverage this significant technological advancement. As we unveil this new chapter, we anticipate fruitful interactions and innovative applications arising from the collective contributions of our user base.

Jamba has emerged as the leading long context model, specifically crafted for builders and tailored to meet enterprise requirements. It outperforms other prominent models of similar scale with its exceptional latency and features a groundbreaking 256k context window, the largest available. Utilizing the innovative Mamba-Transformer MoE architecture, Jamba prioritizes cost efficiency and operational effectiveness. Among its out-of-the-box features are function calls, JSON mode output, document objects, and citation mode, all aimed at improving the overall user experience. The Jamba 1.5 models excel in performance across their expansive context window and consistently achieve top-tier scores on various quality assessment metrics. Enterprises can take advantage of secure deployment options customized to their specific needs, which facilitates seamless integration with existing systems. Furthermore, Jamba is readily accessible via our robust SaaS platform, and deployment options also include collaboration with strategic partners, providing users with added flexibility. For organizations that require specialized solutions, we offer dedicated management and ongoing pre-training services, ensuring that each client can make the most of Jamba’s capabilities. This level of adaptability and support positions Jamba as a premier choice for enterprises in search of innovative and effective solutions for their needs. Additionally, Jamba's commitment to continuous improvement ensures that it remains at the forefront of technological advancements, further solidifying its reputation as a trusted partner for businesses.

Pre-trained language models have advanced significantly, demonstrating exceptional performance in various Natural Language Processing (NLP) tasks. The remarkable features of GPT-3 illustrate that scaling these models can lead to the discovery of their immense capabilities. Recently, the introduction of a comprehensive framework called ERNIE 3.0 has allowed for the pre-training of large-scale models infused with knowledge, resulting in a model with an impressive 10 billion parameters. This version of ERNIE 3.0 has outperformed many leading models across numerous NLP challenges. In our pursuit of exploring the impact of scaling, we have created an even larger model named ERNIE 3.0 Titan, which boasts up to 260 billion parameters and is developed on the PaddlePaddle framework. Moreover, we have incorporated a self-supervised adversarial loss coupled with a controllable language modeling loss, which empowers ERNIE 3.0 Titan to generate text that is both accurate and adaptable, thus extending the limits of what these models can achieve. This innovative methodology not only improves the model's overall performance but also paves the way for new research opportunities in the fields of text generation and fine-tuning control. As the landscape of NLP continues to evolve, the advancements in these models promise to drive further breakthroughs in understanding and generating human language.

The Gemma family is composed of advanced and lightweight models that are built upon the same groundbreaking research and technology as the Gemini line. These state-of-the-art models come with powerful security features that foster responsible and trustworthy AI usage, a result of meticulously selected data sets and comprehensive refinements. Remarkably, the Gemma models perform exceptionally well in their varied sizes—2B, 7B, 9B, and 27B—frequently surpassing the capabilities of some larger open models. With the launch of Keras 3.0, users benefit from seamless integration with JAX, TensorFlow, and PyTorch, allowing for adaptable framework choices tailored to specific tasks. Optimized for peak performance and exceptional efficiency, Gemma 2 in particular is designed for swift inference on a wide range of hardware platforms. Moreover, the Gemma family encompasses a variety of models tailored to meet different use cases, ensuring effective adaptation to user needs. These lightweight language models are equipped with a decoder and have undergone training on a broad spectrum of textual data, programming code, and mathematical concepts, which significantly boosts their versatility and utility across numerous applications. This diverse approach not only enhances their performance but also positions them as a valuable resource for developers and researchers alike.