Top Arcee AI Alternatives

FLUX.2 represents a frontier-level leap in visual intelligence, built to support the demands of modern creative production rather than simple demos. It combines precise prompt following, multi-reference consistency, and coherent world modeling to produce images that adhere to brand rules, layout constraints, and detailed styling instructions. The model excels at everything from photoreal product renders to infographic-grade typography, maintaining clarity and stability even with tightly structured prompts. Its ability to edit and generate at resolutions up to 4 megapixels makes it suitable for advertising, visualization, and enterprise-grade creative pipelines. FLUX.2’s core architecture fuses a large Mistral-3-based vision-language model with a powerful latent rectified-flow transformer, capturing scene structure, spatial relationships, and authentic lighting cues. The rebuilt VAE improves fidelity and learnability while keeping inference efficient—advancing the industry’s understanding of the learnability-quality-compression tradeoff. Developers can choose between FLUX.2 [pro] for top-tier results, FLUX.2 [flex] for parameter-level control, FLUX.2 [dev] for open-weight self-hosting, and FLUX.2 [klein] for a lightweight Apache-licensed option. Each model unifies text-to-image, image editing, and multi-input conditioning in a single architecture. With industry-leading performance and an open-core philosophy, FLUX.2 is positioned to become foundational creative infrastructure across design, research, and enterprise. It also pushes the field closer to multimodal systems that blend perception, memory, and reasoning in an open and transparent way.

Mistral Large 3 is a frontier-scale open AI model built on a sophisticated Mixture-of-Experts framework that unlocks 41B active parameters per step while maintaining a massive 675B total parameter capacity. This architecture lets the model deliver exceptional reasoning, multilingual mastery, and multimodal understanding at a fraction of the compute cost typically associated with models of this scale. Trained entirely from scratch on 3,000 NVIDIA H200 GPUs, it reaches competitive alignment performance with leading closed models, while achieving best-in-class results among permissively licensed alternatives. Mistral Large 3 includes base and instruction editions, supports images natively, and will soon introduce a reasoning-optimized version capable of even deeper thought chains. Its inference stack has been carefully co-designed with NVIDIA, enabling efficient low-precision execution, optimized MoE kernels, speculative decoding, and smooth long-context handling on Blackwell NVL72 systems and enterprise-grade clusters. Through collaborations with vLLM and Red Hat, developers gain an easy path to run Large 3 on single-node 8×A100 or 8×H100 environments with strong throughput and stability. The model is available across Mistral AI Studio, Amazon Bedrock, Azure Foundry, Hugging Face, Fireworks, OpenRouter, Modal, and more, ensuring turnkey access for development teams. Enterprises can go further with Mistral’s custom-training program, tailoring the model to proprietary data, regulatory workflows, or industry-specific tasks. From agentic applications to multilingual customer automation, creative workflows, edge deployment, and advanced tool-use systems, Mistral Large 3 adapts to a wide range of production scenarios. With this release, Mistral positions the 3-series as a complete family—spanning lightweight edge models to frontier-scale MoE intelligence—while remaining fully open, customizable, and performance-optimized across the stack.

GLM-5.1 marks the newest evolution in Z.ai’s GLM lineup, designed as a state-of-the-art AI model focused on agents, specifically for tasks involving coding, logical reasoning, and overseeing long-term processes. This version builds on the foundation set by GLM-5, which utilizes a Mixture-of-Experts (MoE) framework to maximize performance while keeping inference costs low, supporting a broader vision of making weight models available to developers. A key feature of GLM-5.1 is its ability to promote agentic behavior, enabling it to plan, execute, and enhance multi-step tasks rather than just responding to single prompts. The model is meticulously crafted to handle complex workflows, such as troubleshooting code, navigating repositories, and conducting sequential tasks, all while preserving context over extended periods. Compared to earlier models, GLM-5.1 provides improved reliability during prolonged interactions, ensuring consistency throughout longer sessions and reducing errors in multi-step reasoning tasks. Furthermore, this advancement represents a significant step forward in the realm of AI, especially in its proficiency for managing intricate task workflows with ease. With its innovative features, GLM-5.1 sets a new standard for what agent-focused AI can achieve in practical applications.

EXAONE Deep is a suite of sophisticated language models developed by LG AI Research, featuring configurations of 2.4 billion, 7.8 billion, and 32 billion parameters. These models are particularly adept at tackling a range of reasoning tasks, excelling in domains like mathematics and programming evaluations. Notably, the 2.4B variant stands out among its peers of comparable size, while the 7.8B model surpasses both open-weight counterparts and the proprietary model OpenAI o1-mini. Additionally, the 32B variant competes strongly with leading open-weight models in the industry. The accompanying repository not only provides comprehensive documentation, including performance metrics and quick-start guides for utilizing EXAONE Deep models with the Transformers library, but also offers in-depth explanations of quantized EXAONE Deep weights structured in AWQ and GGUF formats. Users will also find instructions on how to operate these models locally using tools like llama.cpp and Ollama, thereby broadening their understanding of the EXAONE Deep models' potential and ensuring easier access to their powerful capabilities. This resource aims to empower users by facilitating a deeper engagement with the advanced functionalities of the models.

Tülu 3 represents a state-of-the-art language model designed by the Allen Institute for AI (Ai2) with the objective of enhancing expertise in various domains such as knowledge, reasoning, mathematics, coding, and safety. Built on the foundation of the Llama 3 Base, it undergoes an intricate four-phase post-training process: meticulous prompt curation and synthesis, supervised fine-tuning across a diverse range of prompts and outputs, preference tuning with both off-policy and on-policy data, and a distinctive reinforcement learning approach that bolsters specific skills through quantifiable rewards. This open-source model is distinguished by its commitment to transparency, providing comprehensive access to its training data, coding resources, and evaluation metrics, thus helping to reduce the performance gap typically seen between open-source and proprietary fine-tuning methodologies. Performance evaluations indicate that Tülu 3 excels beyond similarly sized models, such as Llama 3.1-Instruct and Qwen2.5-Instruct, across multiple benchmarks, emphasizing its superior effectiveness. The ongoing evolution of Tülu 3 not only underscores a dedication to enhancing AI capabilities but also fosters an inclusive and transparent technological landscape. As such, it paves the way for future advancements in artificial intelligence that prioritize collaboration and accessibility for all users.

Trinity Large Thinking is a cutting-edge open-source reasoning framework developed by Arcee AI, specifically designed for tackling complex, multi-step problems and workflows that involve autonomous agents requiring extensive planning and diverse tool utilization. With an impressive sparse Mixture-of-Experts architecture, it encompasses around 400 billion parameters, activating about 13 billion for each token, which not only boosts its operational efficiency but also fortifies its reasoning capabilities across various tasks, such as mathematical computations, code generation, and thorough analysis. A significant innovation of this model is its capacity for extended chain-of-thought reasoning, enabling it to generate intermediate "thinking traces" prior to presenting final results, which significantly enhances accuracy and dependability in intricate scenarios. Additionally, Trinity Large Thinking supports a generous context window of up to 262K tokens, which empowers it to effectively handle lengthy documents, maintain context during extended interactions, and operate smoothly within continuous agent loops. This exemplary design showcases a firm commitment to advancing the limits of automated reasoning systems, paving the way for more sophisticated applications in the future. As technology evolves, the potential for further enhancements in reasoning models like this one remains vast and exciting.

OLMo 2 is a suite of fully open language models developed by the Allen Institute for AI (AI2), designed to provide researchers and developers with straightforward access to training datasets, open-source code, reproducible training methods, and extensive evaluations. These models are trained on a remarkable dataset consisting of up to 5 trillion tokens and are competitive with leading open-weight models such as Llama 3.1, especially in English academic assessments. A significant emphasis of OLMo 2 lies in maintaining training stability, utilizing techniques to reduce loss spikes during prolonged training sessions, and implementing staged training interventions to address capability weaknesses in the later phases of pretraining. Furthermore, the models incorporate advanced post-training methodologies inspired by AI2's Tülu 3, resulting in the creation of OLMo 2-Instruct models. To support continuous enhancements during the development lifecycle, an actionable evaluation framework called the Open Language Modeling Evaluation System (OLMES) has been established, featuring 20 benchmarks that assess vital capabilities. This thorough methodology not only promotes transparency but also actively encourages improvements in the performance of language models, ensuring they remain at the forefront of AI advancements. Ultimately, OLMo 2 aims to empower the research community by providing resources that foster innovation and collaboration in language modeling.

Qwen3.6 is a next-generation large language model developed by Alibaba, designed to deliver advanced reasoning, coding, and multimodal capabilities. It builds on the Qwen3.5 series with a strong emphasis on stability, efficiency, and real-world usability. The model supports multimodal inputs, enabling it to process text, images, and video for more complex analysis and decision-making. One of its key strengths is agentic AI, allowing it to perform multi-step tasks and operate more autonomously in workflows. Qwen3.6 is particularly optimized for coding, capable of handling complex engineering tasks at a repository level rather than just individual functions. It uses a mixture-of-experts architecture, with billions of parameters but only a subset activated during each inference, improving efficiency. The model is available in both open-weight and proprietary versions, giving developers flexibility in deployment and customization. It can be integrated into enterprise systems, APIs, and cloud environments for production use. Qwen3.6 also offers strong multimodal reasoning, enabling it to analyze documents, visuals, and structured data together. It is designed to support a wide range of applications, from software development to data analysis and automation. The model includes enhancements in performance, scalability, and usability compared to earlier versions. It reflects a broader shift toward agent-based AI systems that can execute tasks rather than just provide responses. Overall, Qwen3.6 represents a powerful and versatile AI model for modern enterprise and developer use cases.

Liquid AI's LFM2.5 marks a significant evolution in on-device AI foundation models, designed to optimize efficiency and performance for AI inference across edge devices, including smartphones, laptops, vehicles, IoT systems, and various embedded hardware, all while eliminating reliance on cloud computing. This upgraded version builds on the previous LFM2 framework by significantly increasing the scale of pretraining and enhancing the stages of reinforcement learning, leading to a collection of hybrid models that feature approximately 1.2 billion parameters and successfully balance adherence to instructions, reasoning capabilities, and multimodal functions for real-world applications. The LFM2.5 lineup includes various models, such as Base (for fine-tuning and personalization), Instruct (tailored for general-purpose instruction), Japanese-optimized, Vision-Language, and Audio-Language editions, all carefully designed for swift on-device inference, even under strict memory constraints. Additionally, these models are offered as open-weight alternatives, enabling easy deployment through platforms like llama.cpp, MLX, vLLM, and ONNX, which enhances flexibility for developers. With these advancements, LFM2.5 not only solidifies its position as a powerful solution for a wide range of AI-driven tasks but also demonstrates Liquid AI's commitment to pushing the boundaries of what is possible with on-device technology. The combination of scalability and versatility ensures that developers can harness the full potential of AI in practical, everyday scenarios.

Phi-4-reasoning-plus is an enhanced reasoning model that boasts 14 billion parameters, significantly improving upon the capabilities of the original Phi-4-reasoning. Utilizing reinforcement learning, it achieves greater inference efficiency by processing 1.5 times the number of tokens that its predecessor could manage, leading to enhanced accuracy in its outputs. Impressively, this model surpasses both OpenAI's o1-mini and DeepSeek-R1 on various benchmarks, tackling complex challenges in mathematical reasoning and high-level scientific questions. In a remarkable feat, it even outshines the much larger DeepSeek-R1, which contains 671 billion parameters, in the esteemed AIME 2025 assessment, a key qualifier for the USA Math Olympiad. Additionally, Phi-4-reasoning-plus is readily available on platforms such as Azure AI Foundry and HuggingFace, streamlining access for developers and researchers eager to utilize its advanced features. Its cutting-edge design not only showcases its capabilities but also establishes it as a formidable player in the competitive landscape of reasoning models. This positions Phi-4-reasoning-plus as a preferred choice for users seeking high-performance reasoning solutions.

Qwen3.5 is an advanced open-weight multimodal AI system built to serve as the foundation for native digital agents capable of reasoning across text, images, and video. The primary release, Qwen3.5-397B-A17B, introduces a hybrid architecture that combines Gated DeltaNet linear attention with a sparse mixture-of-experts design, activating just 17 billion parameters per inference pass while maintaining a total parameter count of 397 billion. This selective activation dramatically improves decoding throughput and cost efficiency without sacrificing benchmark-level performance. Qwen3.5 demonstrates strong results across knowledge, multilingual reasoning, coding, STEM tasks, search agents, visual question answering, document understanding, and spatial intelligence benchmarks. The hosted Qwen3.5-Plus variant offers a default one-million-token context window and integrated tool usage such as web search and code interpretation for adaptive problem-solving. Expanded multilingual support now covers 201 languages and dialects, backed by a 250k vocabulary that enhances encoding and decoding efficiency across global use cases. The model is natively multimodal, using early fusion techniques and large-scale visual-text pretraining to outperform prior Qwen-VL systems in scientific reasoning and video analysis. Infrastructure innovations such as heterogeneous parallel training, FP8 precision pipelines, and disaggregated reinforcement learning frameworks enable near-text baseline throughput even with mixed multimodal inputs. Extensive reinforcement learning across diverse and generalized environments improves long-horizon planning, multi-turn interactions, and tool-augmented workflows. Designed for developers, researchers, and enterprises, Qwen3.5 supports scalable deployment through Alibaba Cloud Model Studio while paving the way toward persistent, economically aware, autonomous AI agents.

MiniMax-M2.1 is a high-performance, open-source agentic language model designed for modern development and automation needs. It was created to challenge the idea that advanced AI agents must remain proprietary. The model is optimized for software engineering, tool usage, and long-horizon reasoning tasks. MiniMax-M2.1 performs strongly in multilingual coding and cross-platform development scenarios. It supports building autonomous agents capable of executing complex, multi-step workflows. Developers can deploy the model locally, ensuring full control over data and execution. The architecture emphasizes robustness, consistency, and instruction accuracy. MiniMax-M2.1 demonstrates competitive results across industry-standard coding and agent benchmarks. It generalizes well across different agent frameworks and inference engines. The model is suitable for full-stack application development, automation, and AI-assisted engineering. Open weights allow experimentation, fine-tuning, and research. MiniMax-M2.1 provides a powerful foundation for the next generation of intelligent agents.

DeepSeek V3.1 emerges as a groundbreaking open-weight large language model, featuring an astounding 685-billion parameters and an extensive 128,000-token context window that enables it to process lengthy documents similar to 400-page novels in a single run. This model encompasses integrated capabilities for conversation, reasoning, and code generation within a unified hybrid framework that effectively blends these varied functionalities. Additionally, V3.1 supports multiple tensor formats, allowing developers to optimize performance across different hardware configurations. Initial benchmark tests indicate impressive outcomes, with a notable score of 71.6% on the Aider coding benchmark, placing it on par with or even outperforming competitors like Claude Opus 4, all while maintaining a significantly lower cost. Launched under an open-source license on Hugging Face with minimal promotion, DeepSeek V3.1 aims to transform the availability of advanced AI solutions, potentially challenging the traditional landscape dominated by proprietary models. The model's innovative features and affordability are likely to attract a diverse array of developers eager to implement state-of-the-art AI technologies in their applications, thus fostering a new wave of creativity and efficiency in the tech industry.

Qwen3-Coder-Next is an open-weight language model designed specifically for coding agents and local development, excelling in complex coding reasoning, proficient tool utilization, and effectively managing long-term programming tasks with exceptional efficiency through a mixture-of-experts framework that balances strong capabilities with a resource-conscious design. This model significantly boosts the coding abilities of software developers, AI system designers, and automated coding systems, enabling them to create, troubleshoot, and understand code with a deep contextual insight while skillfully recovering from execution errors, making it particularly suitable for autonomous coding agents and development-focused applications. Additionally, Qwen3-Coder-Next offers remarkable performance comparable to models with larger parameters but operates with a reduced number of active parameters, making it a cost-effective solution for tackling complex and dynamic programming challenges in both research and production environments. Ultimately, this innovative model is designed to enhance the efficiency and effectiveness of the development process, paving the way for more agile and responsive software creation. Its ability to streamline workflows further underscores its potential to transform how programming tasks are approached and executed.

GLM-5 is Z.ai’s most advanced open-source model to date, purpose-built for complex systems engineering, long-horizon planning, and autonomous agent workflows. Building on the foundation of GLM-4.5, it dramatically scales both total parameters and pre-training data while increasing active parameter efficiency. The integration of DeepSeek Sparse Attention allows GLM-5 to maintain strong long-context reasoning capabilities while reducing deployment costs. To improve post-training performance, Z.ai developed slime, an asynchronous reinforcement learning infrastructure that significantly boosts training throughput and iteration speed. As a result, GLM-5 achieves top-tier performance among open-source models across reasoning, coding, and general agent benchmarks. It demonstrates exceptional strength in long-term operational simulations, including leading results on Vending Bench 2, where it manages a year-long simulated business with strong financial outcomes. In coding evaluations such as SWE-bench and Terminal-Bench 2.0, GLM-5 delivers competitive results that narrow the gap with proprietary frontier systems. The model is fully open-sourced under the MIT License and available through Hugging Face, ModelScope, and Z.ai’s developer platforms. Developers can deploy GLM-5 locally using inference frameworks like vLLM and SGLang, including support for non-NVIDIA hardware through optimization and quantization techniques. Through Z.ai, users can access both Chat Mode for fast interactions and Agent Mode for tool-augmented, multi-step task execution. GLM-5 also enables structured document generation, producing ready-to-use .docx, .pdf, and .xlsx files for business and academic workflows. With compatibility across coding agents and cross-application automation frameworks, GLM-5 moves foundation models from conversational assistants toward full-scale work engines.

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.

Qwen3.6-27B stands as an open-source, dense multimodal language model within the Qwen3.6 lineup, crafted to deliver exceptional capabilities in coding, reasoning, and workflows driven by agents, all while utilizing a streamlined parameter count of 27 billion. This model is distinguished by its performance, often surpassing or closely rivaling larger models on critical benchmarks, especially in tasks that involve agent-based coding. It operates in two distinct modes—thinking and non-thinking—allowing it to adjust the depth of its reasoning and the speed of its responses to align with the specific demands of various tasks. Furthermore, it accommodates a broad range of input formats, which includes text, images, and video, demonstrating its adaptability. As an integral part of the Qwen3.6 series, this model emphasizes practical functionality, reliability, and the boost of developer efficiency, drawing on feedback from the community and the practical needs of real-world applications. Its forward-thinking design not only addresses current user requirements but also foresees future developments in the realm of artificial intelligence, ensuring that it remains relevant and effective over time. Thus, Qwen3.6-27B represents a significant step forward in the evolution of language models, integrating innovative features that enhance user interaction and streamline workflows.

Phi-4-reasoning is a sophisticated transformer model that boasts 14 billion parameters, crafted specifically to address complex reasoning tasks such as mathematics, programming, algorithm design, and strategic decision-making. It achieves this through an extensive supervised fine-tuning process, utilizing curated "teachable" prompts and reasoning examples generated via o3-mini, which allows it to produce detailed reasoning sequences while optimizing computational efficiency during inference. By employing outcome-driven reinforcement learning techniques, Phi-4-reasoning is adept at generating longer reasoning pathways. Its performance is remarkable, exceeding that of much larger open-weight models like DeepSeek-R1-Distill-Llama-70B, and it closely rivals the more comprehensive DeepSeek-R1 model across a range of reasoning tasks. Engineered for environments with constrained computing resources or high latency, this model is refined with synthetic data sourced from DeepSeek-R1, ensuring it provides accurate and methodical solutions to problems. The efficiency with which this model processes intricate tasks makes it an indispensable asset in various computational applications, further enhancing its significance in the field. Its innovative design reflects an ongoing commitment to pushing the boundaries of artificial intelligence capabilities.

Step 3.5 Flash represents a state-of-the-art open-source foundational language model crafted for sophisticated reasoning and agent-like functionality, prioritizing efficiency; it employs a sparse Mixture of Experts (MoE) framework that activates roughly 11 billion of its nearly 196 billion parameters for each token, which ensures both dense intelligence and rapid responsiveness. The architecture includes a 3-way Multi-Token Prediction (MTP-3) system, enabling the generation of hundreds of tokens per second and supporting intricate multi-step reasoning and task execution, while efficiently handling extensive contexts through a hybrid sliding window attention technique that reduces computational stress on large datasets or codebases. Its remarkable capabilities in reasoning, coding, and agentic tasks often rival or exceed those of much larger proprietary models, further enhanced by a scalable reinforcement learning mechanism that promotes ongoing self-improvement. This innovative design not only highlights Step 3.5 Flash's effectiveness but also positions it as a transformative force in the domain of AI language models, indicating its vast potential across a plethora of applications. As such, it stands as a testament to the advancements in AI technology, paving the way for future developments.

Qwen2 is a comprehensive array of advanced language models developed by the Qwen team at Alibaba Cloud. This collection includes various models that range from base to instruction-tuned versions, with parameters from 0.5 billion up to an impressive 72 billion, demonstrating both dense configurations and a Mixture-of-Experts architecture. The Qwen2 lineup is designed to surpass many earlier open-weight models, including its predecessor Qwen1.5, while also competing effectively against proprietary models across several benchmarks in domains such as language understanding, text generation, multilingual capabilities, programming, mathematics, and logical reasoning. Additionally, this cutting-edge series is set to significantly influence the artificial intelligence landscape, providing enhanced functionalities that cater to a wide array of applications. As such, the Qwen2 models not only represent a leap in technological advancement but also pave the way for future innovations in the field.

Sky-T1-32B-Preview represents a groundbreaking open-source reasoning model developed by the NovaSky team at UC Berkeley's Sky Computing Lab. It achieves performance levels similar to those of proprietary models like o1-preview across a range of reasoning and coding tests, all while being created for under $450, emphasizing its potential to provide advanced reasoning skills at a lower cost. Fine-tuned from Qwen2.5-32B-Instruct, this model was trained on a carefully selected dataset of 17,000 examples that cover diverse areas, including mathematics and programming. The training was efficiently completed in a mere 19 hours with the aid of eight H100 GPUs using DeepSpeed Zero-3 offloading technology. Notably, every aspect of this project—spanning data, code, and model weights—is fully open-source, enabling both the academic and open-source communities to not only replicate but also enhance the model's functionalities. Such openness promotes a spirit of collaboration and innovation within the artificial intelligence research and development landscape, inviting contributions from various sectors. Ultimately, this initiative represents a significant step forward in making powerful AI tools more accessible to a wider audience.

gpt-oss-120b is a reasoning model focused solely on text, boasting 120 billion parameters, and is released under the Apache 2.0 license while adhering to OpenAI’s usage policies; it has been developed with contributions from the open-source community and is compatible with the Responses API. This model excels at executing instructions and utilizes various tools, including web searches and Python code execution, which allows for a customizable level of reasoning effort and results in detailed chain-of-thought outputs that can seamlessly fit into different workflows. Although it is constructed to comply with OpenAI's safety policies, its open-weight nature poses a risk, as adept users might modify it to bypass these protections, thereby prompting developers and organizations to implement additional safety measures akin to those of managed models. Assessments reveal that gpt-oss-120b falls short of high performance in specialized fields such as biology, chemistry, or cybersecurity, even after attempts at adversarial fine-tuning. Moreover, its introduction does not represent a substantial advancement in biological capabilities, indicating a cautious stance regarding its use. Consequently, it is advisable for users to stay alert to the potential risks associated with its open-weight attributes, and to consider the implications of its deployment in sensitive environments. As awareness of these factors grows, the community's approach to managing such technologies will evolve and adapt.

MiniMax M2.5 is an advanced frontier model designed to deliver real-world productivity across coding, search, agentic tool use, and high-value office tasks. Built on large-scale reinforcement learning across hundreds of thousands of structured environments, it achieves state-of-the-art results on benchmarks such as SWE-Bench Verified, Multi-SWE-Bench, and BrowseComp. The model demonstrates architect-level planning capabilities, decomposing system requirements before generating full-stack code across more than ten programming languages including Go, Python, Rust, TypeScript, and Java. It supports complex development lifecycles, from initial system design and environment setup to iterative feature development and comprehensive code review. With native serving speeds of up to 100 tokens per second, M2.5 significantly reduces task completion time compared to prior versions. Reinforcement learning enhancements improve token efficiency and reduce redundant reasoning rounds, making agentic workflows faster and more precise. The model is available in both M2.5 and M2.5-Lightning variants, offering identical intelligence with different throughput configurations. Its pricing structure dramatically undercuts other frontier models, enabling continuous deployment at a fraction of traditional costs. M2.5 is fully integrated into MiniMax Agent, where standardized Office Skills allow it to generate formatted Word documents, financial models in Excel, and presentation-ready PowerPoint decks. Users can also create reusable domain-specific “Experts” that combine industry frameworks with Office Skills for structured, professional outputs. Internally, MiniMax reports that M2.5 autonomously completes a significant portion of operational tasks, including a majority of newly committed code. By pairing scalable reinforcement learning, high-speed inference, and ultra-low cost, MiniMax M2.5 positions itself as a production-ready engine for complex agent-driven applications.

We are thrilled to introduce Codestral, our first code generation model. This generative AI system, featuring open weights, is designed explicitly for code generation tasks, allowing developers to effortlessly write and interact with code through a single instruction and completion API endpoint. As it gains expertise in both programming languages and English, Codestral is set to enhance the development of advanced AI applications specifically for software engineers. The model is built on a robust foundation that includes a diverse selection of over 80 programming languages, spanning popular choices like Python, Java, C, C++, JavaScript, and Bash, as well as less common languages such as Swift and Fortran. This broad language support guarantees that developers have the tools they need to address a variety of coding challenges and projects. Furthermore, Codestral’s rich language capabilities enable developers to work with confidence across different coding environments, solidifying its role as an essential resource in the programming community. Ultimately, Codestral stands to revolutionize the way developers approach code generation and project execution.

Pixtral Large is a comprehensive multimodal model developed by Mistral AI, boasting an impressive 124 billion parameters that build upon their earlier Mistral Large 2 framework. The architecture consists of a 123-billion-parameter multimodal decoder paired with a 1-billion-parameter vision encoder, which empowers the model to adeptly interpret diverse content such as documents, graphs, and natural images while maintaining excellent text understanding. Furthermore, Pixtral Large can accommodate a substantial context window of 128,000 tokens, enabling it to process at least 30 high-definition images simultaneously with impressive efficiency. Its performance has been validated through exceptional results in benchmarks like MathVista, DocVQA, and VQAv2, surpassing competitors like GPT-4o and Gemini-1.5 Pro. The model is made available for research and educational use under the Mistral Research License, while also offering a separate Mistral Commercial License for businesses. This dual licensing approach enhances its appeal, making Pixtral Large not only a powerful asset for academic research but also a significant contributor to advancements in commercial applications. As a result, the model stands out as a multifaceted tool capable of driving innovation across various fields.

ERNIE-Image is an innovative text-to-image generation model developed by Baidu, designed to create high-quality visuals with a strong emphasis on following user instructions and providing greater control. It employs a single-stream Diffusion Transformer (DiT) architecture, boasting around 8 billion parameters, which allows it to outperform many other open-weight image generation models while remaining efficient in its operations. The model includes a unique prompt enhancement feature that enriches simple user inputs into more detailed and sophisticated descriptions, significantly improving the overall quality and consistency of the images produced. Its strength lies in its ability to follow complex instructions meticulously, which allows for the accurate representation of text within images, the organization of structured layouts, and the crafting of compositions with multiple elements, making it particularly suitable for projects like posters, comics, and multi-panel designs. In addition, ERNIE-Image supports multilingual prompts in languages such as English, Chinese, and Japanese, broadening its accessibility and applicability across various cultural contexts. This adaptability enables users to explore a wider array of creative possibilities, allowing them to visually articulate their concepts in an assortment of environments. As a result, the model not only serves individual creators but also has the potential to impact various industries by facilitating innovative visual storytelling.

Xiaomi MiMo-V2.5-Pro is a cutting-edge open-source AI model built to handle complex reasoning, coding, and long-horizon tasks with high efficiency. It features a Mixture-of-Experts architecture with over one trillion total parameters and a large active parameter set for optimized performance. The model supports an extended context window of up to one million tokens, enabling it to process large amounts of information in a single workflow. It is designed for advanced agentic capabilities, allowing it to autonomously complete multi-step tasks over extended periods. MiMo-V2.5-Pro has demonstrated strong results in benchmarks related to software engineering, reasoning, and general AI performance. It is capable of building complete applications, optimizing engineering systems, and solving complex technical challenges. The model uses hybrid attention mechanisms to balance performance and efficiency across long contexts. It is also optimized for token efficiency, reducing resource usage while maintaining high-quality outputs. The model can integrate with development tools and frameworks to support real-world use cases. Xiaomi has open-sourced MiMo-V2.5-Pro, providing developers with access to its architecture, weights, and deployment tools. This allows organizations to customize and scale the model for their specific needs. Its ability to handle long workflows makes it suitable for tasks that require sustained reasoning and coordination. By combining scalability, efficiency, and advanced intelligence, MiMo-V2.5-Pro represents a significant advancement in open-source AI technology.

Olmo 3 constitutes an extensive series of open models that include versions with 7 billion and 32 billion parameters, delivering outstanding performance in areas such as base functionality, reasoning, instruction, and reinforcement learning, all while ensuring transparency throughout the development process, including access to raw training datasets, intermediate checkpoints, training scripts, extended context support (with a remarkable window of 65,536 tokens), and provenance tools. The backbone of these models is derived from the Dolma 3 dataset, which encompasses about 9 trillion tokens and employs a thoughtful mixture of web content, scientific research, programming code, and comprehensive documents; this meticulous strategy of pre-training, mid-training, and long-context usage results in base models that receive further refinement through supervised fine-tuning, preference optimization, and reinforcement learning with accountable rewards, leading to the emergence of the Think and Instruct versions. Importantly, the 32 billion Think model has earned recognition as the most formidable fully open reasoning model available thus far, showcasing a performance level that closely competes with that of proprietary models in disciplines such as mathematics, programming, and complex reasoning tasks, highlighting a considerable leap forward in the realm of open model innovation. This breakthrough not only emphasizes the capabilities of open-source models but also suggests a promising future where they can effectively rival conventional closed systems across a range of sophisticated applications, potentially reshaping the landscape of artificial intelligence.

gpt-oss-20b is a robust text-only reasoning model featuring 20 billion parameters, released under the Apache 2.0 license and shaped by OpenAI’s gpt-oss usage guidelines, aimed at simplifying the integration into customized AI workflows via the Responses API without reliance on proprietary systems. It has been meticulously designed to perform exceptionally in following instructions, offering capabilities like adjustable reasoning effort, detailed chain-of-thought outputs, and the option to leverage native tools such as web search and Python execution, which leads to well-structured and coherent responses. Developers must take responsibility for implementing their own deployment safeguards, including input filtering, output monitoring, and compliance with usage policies, to ensure alignment with protective measures typically associated with hosted solutions and to minimize the risk of malicious or unintended actions. Furthermore, its open-weight architecture is particularly advantageous for on-premises or edge deployments, highlighting the significance of control, customization, and transparency to cater to specific user requirements. This flexibility empowers organizations to adapt the model to their distinct needs while upholding a high standard of operational integrity and performance. As a result, gpt-oss-20b not only enhances user experience but also promotes responsible AI usage across various applications.

Tiny Aya is a suite of multilingual language models created by Cohere Labs, designed to deliver powerful and adaptable artificial intelligence capabilities that can operate effectively on local devices like smartphones and laptops, eliminating the necessity for constant cloud connectivity. This pioneering model focuses on improving text understanding and generation across more than 70 languages, with particular emphasis on lower-resource languages that often go overlooked by traditional models. Constructed with an efficient architecture featuring approximately 3.35 billion parameters, Tiny Aya has been optimized for excellent multilingual performance and computational efficiency, making it particularly suitable for use in edge computing environments and offline applications. Additionally, the models are structured to allow for downstream adaptation and instruction tuning, which enables developers to customize the models’ functionalities for various specific applications while maintaining robust performance across different languages. Ultimately, Tiny Aya not only broadens the accessibility of cutting-edge AI technologies but also equips developers with the tools needed to craft tailored applications that cater to a wide array of linguistic requirements, thus fostering greater inclusivity in AI-driven solutions. This capacity for customization ensures that Tiny Aya can evolve alongside the needs of its users, making it a versatile choice in the ever-changing landscape of AI development.