Alternatives to Devstral Small 2

DeepSWE is an innovative and fully open-source coding agent that utilizes the Qwen3-32B foundation model, trained solely through reinforcement learning (RL) without any supervised fine-tuning or reliance on proprietary model distillation. Created with rLLM, which is Agentica’s open-source RL framework for language-based agents, DeepSWE operates as a functional agent within a simulated development environment facilitated by the R2E-Gym framework. This allows it to leverage a variety of tools, including a file editor, search capabilities, shell execution, and submission features, enabling the agent to efficiently navigate codebases, modify multiple files, compile code, run tests, and iteratively create patches or complete complex engineering tasks. Beyond simple code generation, DeepSWE showcases advanced emergent behaviors; when faced with bugs or new feature requests, it thoughtfully reasons through edge cases, searches for existing tests within the codebase, suggests patches, develops additional tests to prevent regressions, and adapts its cognitive approach based on the task at hand. This flexibility and capability make DeepSWE a powerful tool in the realm of software development.

DeepCoder, an entirely open-source model for code reasoning and generation, has been developed through a partnership between Agentica Project and Together AI. Leveraging the foundation of DeepSeek-R1-Distilled-Qwen-14B, it has undergone fine-tuning via distributed reinforcement learning, achieving a notable accuracy of 60.6% on LiveCodeBench, which marks an 8% enhancement over its predecessor. This level of performance rivals that of proprietary models like o3-mini (2025-01-031 Low) and o1, all while operating with only 14 billion parameters. The training process spanned 2.5 weeks on 32 H100 GPUs, utilizing a carefully curated dataset of approximately 24,000 coding challenges sourced from validated platforms, including TACO-Verified, PrimeIntellect SYNTHETIC-1, and submissions to LiveCodeBench. Each problem mandated a legitimate solution along with a minimum of five unit tests to guarantee reliability during reinforcement learning training. Furthermore, to effectively manage long-range context, DeepCoder incorporates strategies such as iterative context lengthening and overlong filtering, ensuring it remains adept at handling complex coding tasks. This innovative approach allows DeepCoder to maintain high standards of accuracy and reliability in its code generation capabilities.

Devstral 2 represents a cutting-edge, open-source AI model designed specifically for software engineering, going beyond mere code suggestion to comprehend and manipulate entire codebases, which allows it to perform tasks such as multi-file modifications, bug corrections, refactoring, dependency management, and generating context-aware code. The Devstral 2 suite comprises a robust 123-billion-parameter model and a more compact 24-billion-parameter version, known as “Devstral Small 2,” providing teams with the adaptability they need; the larger variant is optimized for complex coding challenges that require a thorough understanding of context, while the smaller version is suitable for operation on less powerful hardware. With an impressive context window of up to 256 K tokens, Devstral 2 can analyze large repositories, monitor project histories, and ensure a coherent grasp of extensive files, which is particularly beneficial for tackling the complexities of real-world projects. The command-line interface (CLI) enhances the model's capabilities by keeping track of project metadata, Git statuses, and the directory structure, thereby enriching the context for the AI and rendering “vibe-coding” even more effective. This combination of advanced features positions Devstral 2 as a transformative tool in the software development landscape.

The Nemotron 3 Nano stands out as the tiniest model within NVIDIA's Nemotron 3 lineup, specifically designed for agentic AI tasks that require robust reasoning and conversational skills while maintaining cost-effective inference. This hybrid Mamba-Transformer Mixture-of-Experts model boasts 3.2 billion active parameters, 3.6 billion when including embeddings, and a total of 31.6 billion parameters. NVIDIA asserts that this model offers greater accuracy compared to its predecessor, the Nemotron 2 Nano, all while utilizing less than half of the parameters during each forward pass, thus enhancing efficiency without compromising on performance. It is also claimed to surpass the accuracy of both GPT-OSS-20B and Qwen3-30B-A3B-Thinking-2507 across various widely-used benchmarks. With an 8K input and 16K output setting utilizing a single H200, the model achieves an inference throughput that is 3.3 times greater than that of Qwen3-30B-A3B and 2.2 times that of GPT-OSS-20B. Additionally, the Nemotron 3 Nano is capable of handling context lengths of up to 1 million tokens, further establishing its superiority over GPT-OSS-20B and Qwen3-30B-A3B-Instruct-2507. This remarkable combination of features positions it as a leading choice for advanced AI applications that demand both precision and efficiency.

GPT-5.2-Codex is a next-generation coding model created to support advanced, agent-driven software development. Built on the GPT-5.2 architecture, it is fine-tuned specifically for real-world engineering tasks. The model excels at working across large codebases while preserving context over long sessions. It handles complex refactors, migrations, and multi-step implementations more reliably than previous Codex models. GPT-5.2-Codex demonstrates top-tier performance in realistic terminal environments. Enhanced tool-calling and improved factual accuracy make it suitable for production workflows. The model is also significantly stronger in cybersecurity-related tasks. It can assist with vulnerability research and defensive security analysis. GPT-5.2-Codex includes safeguards designed to support responsible deployment. It represents a major advancement in professional-grade coding AI.

Devstral is a collaborative effort between Mistral AI and All Hands AI, resulting in an open-source large language model specifically tailored for software engineering. This model demonstrates remarkable proficiency in navigating intricate codebases, managing edits across numerous files, and addressing practical problems, achieving a notable score of 46.8% on the SWE-Bench Verified benchmark, which is superior to all other open-source models. Based on Mistral-Small-3.1, Devstral boasts an extensive context window supporting up to 128,000 tokens. It is designed for optimal performance on high-performance hardware setups, such as Macs equipped with 32GB of RAM or Nvidia RTX 4090 GPUs, and supports various inference frameworks including vLLM, Transformers, and Ollama. Released under the Apache 2.0 license, Devstral is freely accessible on platforms like Hugging Face, Ollama, Kaggle, Unsloth, and LM Studio, allowing developers to integrate its capabilities into their projects seamlessly. This model not only enhances productivity for software engineers but also serves as a valuable resource for anyone working with code.

Qwen3-Coder is a versatile coding model that comes in various sizes, prominently featuring the 480B-parameter Mixture-of-Experts version with 35B active parameters, which naturally accommodates 256K-token contexts that can be extended to 1M tokens. This model achieves impressive performance that rivals Claude Sonnet 4, having undergone pre-training on 7.5 trillion tokens, with 70% of that being code, and utilizing synthetic data refined through Qwen2.5-Coder to enhance both coding skills and overall capabilities. Furthermore, the model benefits from post-training techniques that leverage extensive, execution-guided reinforcement learning, which facilitates the generation of diverse test cases across 20,000 parallel environments, thereby excelling in multi-turn software engineering tasks such as SWE-Bench Verified without needing test-time scaling. In addition to the model itself, the open-source Qwen Code CLI, derived from Gemini Code, empowers users to deploy Qwen3-Coder in dynamic workflows with tailored prompts and function calling protocols, while also offering smooth integration with Node.js, OpenAI SDKs, and environment variables. This comprehensive ecosystem supports developers in optimizing their coding projects effectively and efficiently.

Stable LM represents a significant advancement in the field of language models by leveraging our previous experience with open-source initiatives, particularly in collaboration with EleutherAI, a nonprofit research organization. This journey includes the development of notable models such as GPT-J, GPT-NeoX, and the Pythia suite, all of which were trained on The Pile open-source dataset, while many contemporary open-source models like Cerebras-GPT and Dolly-2 have drawn inspiration from this foundational work. Unlike its predecessors, Stable LM is trained on an innovative dataset that is three times the size of The Pile, encompassing a staggering 1.5 trillion tokens. We plan to share more information about this dataset in the near future. The extensive nature of this dataset enables Stable LM to excel remarkably in both conversational and coding scenarios, despite its relatively modest size of 3 to 7 billion parameters when compared to larger models like GPT-3, which boasts 175 billion parameters. Designed for versatility, Stable LM 3B is a streamlined model that can efficiently function on portable devices such as laptops and handheld gadgets, making us enthusiastic about its practical applications and mobility. Overall, the development of Stable LM marks a pivotal step towards creating more efficient and accessible language models for a wider audience.

GLM-4.1V is an advanced vision-language model that offers a robust and streamlined multimodal capability for reasoning and understanding across various forms of media, including images, text, and documents. The 9-billion-parameter version, known as GLM-4.1V-9B-Thinking, is developed on the foundation of GLM-4-9B and has been improved through a unique training approach that employs Reinforcement Learning with Curriculum Sampling (RLCS). This model accommodates a context window of 64k tokens and can process high-resolution inputs, supporting images up to 4K resolution with any aspect ratio, which allows it to tackle intricate tasks such as optical character recognition, image captioning, chart and document parsing, video analysis, scene comprehension, and GUI-agent workflows, including the interpretation of screenshots and recognition of UI elements. In benchmark tests conducted at the 10 B-parameter scale, GLM-4.1V-9B-Thinking demonstrated exceptional capabilities, achieving the highest performance on 23 out of 28 evaluated tasks. Its advancements signify a substantial leap forward in the integration of visual and textual data, setting a new standard for multimodal models in various applications.

Anthropic’s release of Claude Opus 4.5 introduces a frontier AI model that excels at coding, complex reasoning, deep research, and long-context tasks. It sets new performance records on real-world engineering benchmarks, handling multi-system debugging, ambiguous instructions, and cross-domain problem solving with greater precision than earlier versions. Testers and early customers reported that Opus 4.5 “just gets it,” offering creative reasoning strategies that even benchmarks fail to anticipate. Beyond raw capability, the model brings stronger alignment and safety, with notable advances in prompt-injection resistance and behavior consistency in high-stakes scenarios. The Claude Developer Platform also gains richer controls including effort tuning, multi-agent orchestration, and context management improvements that significantly boost efficiency. Claude Code becomes more powerful with enhanced planning abilities, multi-session desktop support, and better execution of complex development workflows. In the Claude apps, extended memory and automatic context summarization enable longer, uninterrupted conversations. Together, these upgrades showcase Opus 4.5 as a highly capable, secure, and versatile model designed for both professional workloads and everyday use.

Mistral 3 represents the newest iteration of open-weight AI models developed by Mistral AI, encompassing a diverse range of models that span from compact, edge-optimized versions to a leading large-scale multimodal model. This lineup features three efficient “Ministral 3” models with 3 billion, 8 billion, and 14 billion parameters, tailored for deployment on devices with limited resources, such as laptops, drones, or other edge devices. Additionally, there is the robust “Mistral Large 3,” which is a sparse mixture-of-experts model boasting a staggering 675 billion total parameters, with 41 billion of them being active. These models are designed to handle multimodal and multilingual tasks, excelling not only in text processing but also in image comprehension, and they have showcased exceptional performance on general queries, multilingual dialogues, and multimodal inputs. Furthermore, both the base and instruction-fine-tuned versions are made available under the Apache 2.0 license, allowing for extensive customization and integration into various enterprise and open-source initiatives. This flexibility in licensing encourages innovation and collaboration among developers and organizations alike.

Upstage has unveiled Solar Pro 2, a cutting-edge large language model designed for frontier-scale applications, capable of managing intricate tasks and workflows in various sectors including finance, healthcare, and law. This model is built on a streamlined architecture with 31 billion parameters, ensuring exceptional multilingual capabilities, particularly in Korean, where it surpasses even larger models on key benchmarks such as Ko-MMLU, Hae-Rae, and Ko-IFEval, while maintaining strong performance in English and Japanese as well. In addition to its advanced language comprehension and generation abilities, Solar Pro 2 incorporates a sophisticated Reasoning Mode that significantly enhances the accuracy of multi-step tasks across a wide array of challenges, from general reasoning assessments (MMLU, MMLU-Pro, HumanEval) to intricate mathematics problems (Math500, AIME) and software engineering tasks (SWE-Bench Agentless), achieving problem-solving efficiency that rivals or even surpasses that of models with double the parameters. Furthermore, its enhanced tool-use capabilities allow the model to effectively engage with external APIs and data, broadening its applicability in real-world scenarios. This innovative design not only demonstrates exceptional versatility but also positions Solar Pro 2 as a formidable player in the evolving landscape of AI technologies.

Reka Flash 3 is a cutting-edge multimodal AI model with 21 billion parameters, crafted by Reka AI to perform exceptionally well in tasks such as general conversation, coding, following instructions, and executing functions. This model adeptly handles and analyzes a myriad of inputs, including text, images, video, and audio, providing a versatile and compact solution for a wide range of applications. Built from the ground up, Reka Flash 3 was trained on a rich array of datasets, encompassing both publicly available and synthetic information, and it underwent a meticulous instruction tuning process with high-quality selected data to fine-tune its capabilities. The final phase of its training involved employing reinforcement learning techniques, specifically using the REINFORCE Leave One-Out (RLOO) method, which combined both model-based and rule-based rewards to significantly improve its reasoning skills. With an impressive context length of 32,000 tokens, Reka Flash 3 competes effectively with proprietary models like OpenAI's o1-mini, making it an excellent choice for applications requiring low latency or on-device processing. The model operates at full precision with a memory requirement of 39GB (fp16), although it can be efficiently reduced to just 11GB through the use of 4-bit quantization, demonstrating its adaptability for various deployment scenarios. Overall, Reka Flash 3 represents a significant advancement in multimodal AI technology, capable of meeting diverse user needs across multiple platforms.

Magistral is the inaugural language model family from Mistral AI that emphasizes reasoning, offered in two variants: Magistral Small, a 24 billion parameter open-weight model accessible under Apache 2.0 via Hugging Face, and Magistral Medium, a more robust enterprise-grade version that can be accessed through Mistral's API, the Le Chat platform, and various major cloud marketplaces. Designed for specific domains, it excels in transparent, multilingual reasoning across diverse tasks such as mathematics, physics, structured calculations, programmatic logic, decision trees, and rule-based systems, generating outputs that follow a chain of thought in the user's preferred language, which can be easily tracked and validated. This release signifies a transition towards more compact yet highly effective transparent AI reasoning capabilities. Currently, Magistral Medium is in preview on platforms including Le Chat, the API, SageMaker, WatsonX, Azure AI, and Google Cloud Marketplace. Its design is particularly suited for general-purpose applications that necessitate extended thought processes and improved accuracy compared to traditional non-reasoning language models. The introduction of Magistral represents a significant advancement in the pursuit of sophisticated reasoning in AI applications.

EXAONE Deep represents a collection of advanced language models that are enhanced for reasoning, created by LG AI Research, and come in sizes of 2.4 billion, 7.8 billion, and 32 billion parameters. These models excel in a variety of reasoning challenges, particularly in areas such as mathematics and coding assessments. Significantly, the EXAONE Deep 2.4B model outshines other models of its size, while the 7.8B variant outperforms both open-weight models of similar dimensions and the proprietary reasoning model known as OpenAI o1-mini. Furthermore, the EXAONE Deep 32B model competes effectively with top-tier open-weight models in the field. The accompanying repository offers extensive documentation that includes performance assessments, quick-start guides for leveraging EXAONE Deep models with the Transformers library, detailed explanations of quantized EXAONE Deep weights formatted in AWQ and GGUF, as well as guidance on how to run these models locally through platforms like llama.cpp and Ollama. Additionally, this resource serves to enhance user understanding and accessibility to the capabilities of EXAONE Deep models.

Introducing Mistral NeMo, our latest and most advanced small model yet, featuring a cutting-edge 12 billion parameters and an expansive context length of 128,000 tokens, all released under the Apache 2.0 license. Developed in partnership with NVIDIA, Mistral NeMo excels in reasoning, world knowledge, and coding proficiency within its category. Its architecture adheres to industry standards, making it user-friendly and a seamless alternative for systems currently utilizing Mistral 7B. To facilitate widespread adoption among researchers and businesses, we have made available both pre-trained base and instruction-tuned checkpoints under the same Apache license. Notably, Mistral NeMo incorporates quantization awareness, allowing for FP8 inference without compromising performance. The model is also tailored for diverse global applications, adept in function calling and boasting a substantial context window. When compared to Mistral 7B, Mistral NeMo significantly outperforms in understanding and executing detailed instructions, showcasing enhanced reasoning skills and the ability to manage complex multi-turn conversations. Moreover, its design positions it as a strong contender for multi-lingual tasks, ensuring versatility across various use cases.

Qwen3-Coder-Next is a language model with open weights, crafted for coding agents and local development, which excels in advanced coding reasoning, adept tool usage, and effective handling of long-term programming challenges with remarkable efficiency, utilizing a mixture-of-experts framework that harmonizes robust capabilities with a resource-efficient approach. This model enhances the coding prowess of software developers, AI system architects, and automated coding processes, allowing them to generate, debug, and comprehend code with a profound contextual grasp while adeptly recovering from execution errors, rendering it ideal for autonomous coding agents and applications focused on development. Furthermore, Qwen3-Coder-Next achieves impressive performance on par with larger parameter models, but does so while consuming fewer active parameters, thus facilitating economical deployment for intricate and evolving programming tasks in both research and production settings, ultimately contributing to a more streamlined development process.

Composer 2 is a high-performance AI coding model available within Cursor, built to handle complex programming tasks with improved accuracy and efficiency. It is trained through advanced pretraining and reinforcement learning, allowing it to solve long-horizon coding problems that involve multiple steps and decisions. The model shows significant improvements across major benchmarks such as Terminal-Bench and SWE-bench Multilingual, reflecting its strong real-world coding capabilities. It delivers faster performance while maintaining high-quality outputs, making it suitable for demanding development workflows. Composer 2 is designed to balance intelligence and cost, offering competitive pricing compared to other frontier models. It also includes a faster variant that provides the same level of intelligence with optimized speed for time-sensitive tasks. The model is integrated directly into the Cursor platform, enabling seamless use within development environments. Its ability to handle complex coding scenarios makes it valuable for both individual developers and teams. Overall, Composer 2 enhances productivity by automating and accelerating software development tasks.

Pixtral Large is an expansive multimodal model featuring 124 billion parameters, crafted by Mistral AI and enhancing their previous Mistral Large 2 framework. This model combines a 123-billion-parameter multimodal decoder with a 1-billion-parameter vision encoder, allowing it to excel in the interpretation of various content types, including documents, charts, and natural images, all while retaining superior text comprehension abilities. With the capability to manage a context window of 128,000 tokens, Pixtral Large can efficiently analyze at least 30 high-resolution images at once. It has achieved remarkable results on benchmarks like MathVista, DocVQA, and VQAv2, outpacing competitors such as GPT-4o and Gemini-1.5 Pro. Available for research and educational purposes under the Mistral Research License, it also has a Mistral Commercial License for business applications. This versatility makes Pixtral Large a valuable tool for both academic research and commercial innovations.

Mistral Large 3 pushes open-source AI into frontier territory with a massive sparse MoE architecture that activates 41B parameters per token while maintaining a highly efficient 675B total parameter design. It sets a new performance standard by combining long-context reasoning, multilingual fluency across 40+ languages, and robust multimodal comprehension within a single unified model. Trained end-to-end on thousands of NVIDIA H200 GPUs, it reaches parity with top closed-source instruction models while remaining fully accessible under the Apache 2.0 license. Developers benefit from optimized deployments through partnerships with NVIDIA, Red Hat, and vLLM, enabling smooth inference on A100, H100, and Blackwell-class systems. The model ships in both base and instruct variants, with a reasoning-enhanced version on the way for even deeper analytical capabilities. Beyond general intelligence, Mistral Large 3 is engineered for enterprise customization, allowing organizations to refine the model on internal datasets or domain-specific tasks. Its efficient token generation and powerful multimodal stack make it ideal for coding, document analysis, knowledge workflows, agentic systems, and multilingual communications. With Mistral Large 3, organizations can finally deploy frontier-class intelligence with full transparency, flexibility, and control.

Qwen3.6 is an advanced AI model from Alibaba that builds on previous Qwen releases with a focus on real-world utility and performance. It is designed as a multimodal large language model capable of understanding and generating text while also processing visual and structured data. The model is optimized for coding tasks, enabling developers to handle complex, repository-level programming workflows. Qwen3.6 uses a mixture-of-experts (MoE) architecture, which activates only a portion of its parameters during inference to improve efficiency. This design allows it to deliver strong performance while reducing computational costs. It is available in both proprietary and open-weight versions, giving developers flexibility in deployment. The model supports integration into enterprise systems and cloud platforms, particularly within Alibaba’s ecosystem. Qwen3.6 also introduces stronger agentic capabilities, allowing it to perform multi-step reasoning and more autonomous task execution. It is designed to handle complex workflows, including engineering, analysis, and decision-making tasks. The model emphasizes stability and responsiveness based on developer feedback. Overall, Qwen3.6 provides a scalable and efficient AI solution for coding, automation, and multimodal applications.

GLM-4.7 is a next-generation AI model built to serve as a powerful coding and reasoning partner. It improves significantly on its predecessor across software engineering, multilingual coding, and terminal interaction benchmarks. GLM-4.7 introduces enhanced agentic behavior by thinking before tool use or execution, improving reliability in long and complex tasks. The model demonstrates strong performance in real-world coding environments and popular coding agents. GLM-4.7 also advances visual and frontend generation, producing modern UI designs and well-structured presentation slides. Its improved tool-use capabilities allow it to browse, analyze, and interact with external systems more effectively. Mathematical and logical reasoning have been strengthened through higher benchmark performance on challenging exams. The model supports flexible reasoning modes, allowing users to trade latency for accuracy. GLM-4.7 can be accessed via Z.ai, OpenRouter, and agent-based coding tools. It is designed for developers who need high performance without excessive cost.

GLM-4.5V-Flash is a vision-language model that is open source and specifically crafted to integrate robust multimodal functionalities into a compact and easily deployable framework. It accommodates various types of inputs including images, videos, documents, and graphical user interfaces, facilitating a range of tasks such as understanding scenes, parsing charts and documents, reading screens, and analyzing multiple images. In contrast to its larger counterparts, GLM-4.5V-Flash maintains a smaller footprint while still embodying essential visual language model features such as visual reasoning, video comprehension, handling GUI tasks, and parsing complex documents. This model can be utilized within “GUI agent” workflows, allowing it to interpret screenshots or desktop captures, identify icons or UI components, and assist with both automated desktop and web tasks. While it may not achieve the performance enhancements seen in the largest models, GLM-4.5V-Flash is highly adaptable for practical multimodal applications where efficiency, reduced resource requirements, and extensive modality support are key considerations. Its design ensures that users can harness powerful functionalities without sacrificing speed or accessibility.

Mistral Medium 3 is an innovative AI model designed to offer high performance at a significantly lower cost, making it an attractive solution for enterprises. It integrates seamlessly with both on-premises and cloud environments, supporting hybrid deployments for more flexibility. This model stands out in professional use cases such as coding, STEM tasks, and multimodal understanding, where it achieves near-competitive results against larger, more expensive models. Additionally, Mistral Medium 3 allows businesses to deploy custom post-training and integrate it into existing systems, making it adaptable to various industry needs. With its impressive performance in coding tasks and real-world human evaluations, Mistral Medium 3 is a cost-effective solution that enables companies to implement AI into their workflows. Its enterprise-focused features, including continuous pretraining and domain-specific fine-tuning, make it a reliable tool for sectors like healthcare, financial services, and energy.

GPT-4.1 nano is a lightweight and fast version of GPT-4.1, designed for applications that prioritize speed and affordability. This model can handle up to 1 million tokens of context, making it suitable for tasks such as text classification, autocompletion, and real-time decision-making. With reduced latency and operational costs, GPT-4.1 nano is the ideal choice for businesses seeking powerful AI capabilities on a budget, without sacrificing essential performance features.

Claude Sonnet 4 is an advanced AI model that enhances coding, reasoning, and problem-solving capabilities, perfect for developers and businesses in need of reliable AI support. This new version of Claude Sonnet significantly improves its predecessor’s capabilities by excelling in coding tasks and delivering precise, clear reasoning. With a 72.7% score on SWE-bench, it offers exceptional performance in software development, app creation, and problem-solving. Claude Sonnet 4’s improved handling of complex instructions and reduced errors in codebase navigation make it the go-to choice for enhancing productivity in technical workflows and software projects.

The o1-mini from OpenAI is an innovative and budget-friendly AI model that specializes in improved reasoning capabilities, especially in STEM areas such as mathematics and programming. As a member of the o1 series, it aims to tackle intricate challenges by allocating more time to analyze and contemplate solutions. Although it is smaller in size and costs 80% less than its counterpart, the o1-preview, the o1-mini remains highly effective in both coding assignments and mathematical reasoning. This makes it an appealing choice for developers and businesses that seek efficient and reliable AI solutions. Furthermore, its affordability does not compromise its performance, allowing a wider range of users to benefit from advanced AI technologies.

Claude Opus 4.1 represents a notable incremental enhancement over its predecessor, Claude Opus 4, designed to elevate coding, agentic reasoning, and data-analysis capabilities while maintaining the same level of deployment complexity. This version boosts coding accuracy to an impressive 74.5 percent on SWE-bench Verified and enhances the depth of research and detailed tracking for agentic search tasks. Furthermore, GitHub has reported significant advancements in multi-file code refactoring, and Rakuten Group emphasizes its ability to accurately identify precise corrections within extensive codebases without introducing any bugs. Independent benchmarks indicate that junior developer test performance has improved by approximately one standard deviation compared to Opus 4, reflecting substantial progress consistent with previous Claude releases.

DeepScaleR is a sophisticated language model comprising 1.5 billion parameters, refined from DeepSeek-R1-Distilled-Qwen-1.5B through the use of distributed reinforcement learning combined with an innovative strategy that incrementally expands its context window from 8,000 to 24,000 tokens during the training process. This model was developed using approximately 40,000 meticulously selected mathematical problems sourced from high-level competition datasets, including AIME (1984–2023), AMC (pre-2023), Omni-MATH, and STILL. Achieving an impressive 43.1% accuracy on the AIME 2024 exam, DeepScaleR demonstrates a significant enhancement of around 14.3 percentage points compared to its base model, and it even outperforms the proprietary O1-Preview model, which is considerably larger. Additionally, it excels on a variety of mathematical benchmarks such as MATH-500, AMC 2023, Minerva Math, and OlympiadBench, indicating that smaller, optimized models fine-tuned with reinforcement learning can rival or surpass the capabilities of larger models in complex reasoning tasks. This advancement underscores the potential of efficient modeling approaches in the realm of mathematical problem-solving.

Qwen2 represents a collection of extensive language models crafted by the Qwen team at Alibaba Cloud. This series encompasses a variety of models, including base and instruction-tuned versions, with parameters varying from 0.5 billion to an impressive 72 billion, showcasing both dense configurations and a Mixture-of-Experts approach. The Qwen2 series aims to outperform many earlier open-weight models, including its predecessor Qwen1.5, while also striving to hold its own against proprietary models across numerous benchmarks in areas such as language comprehension, generation, multilingual functionality, programming, mathematics, and logical reasoning. Furthermore, this innovative series is poised to make a significant impact in the field of artificial intelligence, offering enhanced capabilities for a diverse range of applications.

Tülu 3 is a cutting-edge language model created by the Allen Institute for AI (Ai2) that aims to improve proficiency in fields like knowledge, reasoning, mathematics, coding, and safety. It is based on the Llama 3 Base and undergoes a detailed four-stage post-training regimen: careful prompt curation and synthesis, supervised fine-tuning on a wide array of prompts and completions, preference tuning utilizing both off- and on-policy data, and a unique reinforcement learning strategy that enhances targeted skills through measurable rewards. Notably, this open-source model sets itself apart by ensuring complete transparency, offering access to its training data, code, and evaluation tools, thus bridging the performance divide between open and proprietary fine-tuning techniques. Performance assessments reveal that Tülu 3 surpasses other models with comparable sizes, like Llama 3.1-Instruct and Qwen2.5-Instruct, across an array of benchmarks, highlighting its effectiveness. The continuous development of Tülu 3 signifies the commitment to advancing AI capabilities while promoting an open and accessible approach to technology.

Qwen3.6-27B is an open-source, dense multimodal language model from the Qwen3.6 series, engineered to provide top-tier performance in areas such as coding, reasoning, and agent-driven workflows, all while maintaining an efficient parameter count of 27 billion. This model is recognized for its ability to outperform or compete closely with much larger counterparts on essential benchmarks, particularly excelling in agent-based coding tasks. It features dual operational modes—thinking and non-thinking—that enable it to effectively adapt its reasoning depth and response speed based on the specific requirements of each task. Additionally, it supports a variety of input types, including text, images, and video, showcasing its versatility. As part of the Qwen3.6 lineup, this model prioritizes practical usability, consistency, and the enhancement of developer productivity, reflecting advancements inspired by community insights and real-world application demands. Its innovative design not only responds to immediate user needs but also anticipates future trends in AI development.

MiniMax M2.5 is a next-generation foundation model built to power complex, economically valuable tasks with speed and cost efficiency. Trained using large-scale reinforcement learning across hundreds of thousands of real-world task environments, it excels in coding, tool use, search, and professional office workflows. In programming benchmarks such as SWE-Bench Verified and Multi-SWE-Bench, M2.5 reaches state-of-the-art levels while demonstrating improved multilingual coding performance. The model exhibits architect-level reasoning, planning system structure and feature decomposition before writing code. With throughput speeds of up to 100 tokens per second, it completes complex evaluations significantly faster than earlier versions. Reinforcement learning optimizations enable more precise search rounds and fewer reasoning steps, improving overall efficiency. M2.5 is available in two variants—standard and Lightning—offering identical capabilities with different speed configurations. Pricing is designed to be dramatically lower than competing frontier models, reducing cost barriers for large-scale agent deployment. Integrated into MiniMax Agent, the model supports advanced office skills including Word formatting, Excel financial modeling, and PowerPoint editing. By combining high performance, efficiency, and affordability, MiniMax M2.5 aims to make agent-powered productivity accessible at scale.

This repository showcases the research preview of LongLLaMA, an advanced large language model that can manage extensive contexts of up to 256,000 tokens or potentially more. LongLLaMA is developed on the OpenLLaMA framework and has been fine-tuned utilizing the Focused Transformer (FoT) technique. The underlying code for LongLLaMA is derived from Code Llama. We are releasing a smaller 3B base variant of the LongLLaMA model, which is not instruction-tuned, under an open license (Apache 2.0), along with inference code that accommodates longer contexts available on Hugging Face. This model's weights can seamlessly replace LLaMA in existing systems designed for shorter contexts, specifically those handling up to 2048 tokens. Furthermore, we include evaluation results along with comparisons to the original OpenLLaMA models, thereby providing a comprehensive overview of LongLLaMA's capabilities in the realm of long-context processing.

We are excited to announce the launch of Phi-2, a language model featuring 2.7 billion parameters that excels in reasoning and language comprehension, achieving top-tier results compared to other base models with fewer than 13 billion parameters. In challenging benchmarks, Phi-2 competes with and often surpasses models that are up to 25 times its size, a feat made possible by advancements in model scaling and meticulous curation of training data. Due to its efficient design, Phi-2 serves as an excellent resource for researchers interested in areas such as mechanistic interpretability, enhancing safety measures, or conducting fine-tuning experiments across a broad spectrum of tasks. To promote further exploration and innovation in language modeling, Phi-2 has been integrated into the Azure AI Studio model catalog, encouraging collaboration and development within the research community. Researchers can leverage this model to unlock new insights and push the boundaries of language technology.

Molmo 2 represents a cutting-edge suite of open vision-language models that come with completely accessible weights, training data, and code, thereby advancing the original Molmo series' capabilities in grounded image comprehension to encompass video and multiple image inputs. This evolution enables sophisticated video analysis, including pointing, tracking, dense captioning, and question-answering functionalities, all of which demonstrate robust spatial and temporal reasoning across frames. The suite consists of three distinct models: an 8 billion-parameter variant tailored for comprehensive video grounding and QA tasks, a 4 billion-parameter model that prioritizes efficiency, and a 7 billion-parameter model backed by Olmo, which features a fully open end-to-end architecture that includes the foundational language model. Notably, these new models surpass their predecessors on key benchmarks, setting unprecedented standards for open-model performance in image and video comprehension tasks. Furthermore, they often rival significantly larger proprietary systems while being trained on a much smaller dataset compared to similar closed models, showcasing their efficiency and effectiveness in the field. This impressive achievement marks a significant advancement in the accessibility and performance of AI-driven visual understanding technologies.

Mistral 7B is a language model with 7.3 billion parameters that demonstrates superior performance compared to larger models such as Llama 2 13B on a variety of benchmarks. It utilizes innovative techniques like Grouped-Query Attention (GQA) for improved inference speed and Sliding Window Attention (SWA) to manage lengthy sequences efficiently. Released under the Apache 2.0 license, Mistral 7B is readily available for deployment on different platforms, including both local setups and prominent cloud services. Furthermore, a specialized variant known as Mistral 7B Instruct has shown remarkable capabilities in following instructions, outperforming competitors like Llama 2 13B Chat in specific tasks. This versatility makes Mistral 7B an attractive option for developers and researchers alike.

Claude Sonnet 4.5 represents Anthropic's latest advancement in AI, crafted to thrive in extended coding environments, complex workflows, and heavy computational tasks while prioritizing safety and alignment. It sets new benchmarks with its top-tier performance on the SWE-bench Verified benchmark for software engineering and excels in the OSWorld benchmark for computer usage, demonstrating an impressive capacity to maintain concentration for over 30 hours on intricate, multi-step assignments. Enhancements in tool management, memory capabilities, and context interpretation empower the model to engage in more advanced reasoning, leading to a better grasp of various fields, including finance, law, and STEM, as well as a deeper understanding of coding intricacies. The system incorporates features for context editing and memory management, facilitating prolonged dialogues or multi-agent collaborations, while it also permits code execution and the generation of files within Claude applications. Deployed at AI Safety Level 3 (ASL-3), Sonnet 4.5 is equipped with classifiers that guard against inputs or outputs related to hazardous domains and includes defenses against prompt injection, ensuring a more secure interaction. This model signifies a significant leap forward in the intelligent automation of complex tasks, aiming to reshape how users engage with AI technologies.

We are excited to unveil Codestral, our inaugural code generation model. This open-weight generative AI system is specifically crafted for tasks related to code generation, enabling developers to seamlessly write and engage with code via a unified instruction and completion API endpoint. As it becomes proficient in both programming languages and English, Codestral is poised to facilitate the creation of sophisticated AI applications tailored for software developers. With a training foundation that encompasses a wide array of over 80 programming languages—ranging from widely-used options like Python, Java, C, C++, JavaScript, and Bash to more niche languages such as Swift and Fortran—Codestral ensures a versatile support system for developers tackling various coding challenges and projects. Its extensive language capabilities empower developers to confidently navigate different coding environments, making Codestral an invaluable asset in the programming landscape.

Kimi K2 represents a cutting-edge series of open-source large language models utilizing a mixture-of-experts (MoE) architecture, with a staggering 1 trillion parameters in total and 32 billion activated parameters tailored for optimized task execution. Utilizing the Muon optimizer, it has been trained on a substantial dataset of over 15.5 trillion tokens, with its performance enhanced by MuonClip’s attention-logit clamping mechanism, resulting in remarkable capabilities in areas such as advanced knowledge comprehension, logical reasoning, mathematics, programming, and various agentic operations. Moonshot AI offers two distinct versions: Kimi-K2-Base, designed for research-level fine-tuning, and Kimi-K2-Instruct, which is pre-trained for immediate applications in chat and tool interactions, facilitating both customized development and seamless integration of agentic features. Comparative benchmarks indicate that Kimi K2 surpasses other leading open-source models and competes effectively with top proprietary systems, particularly excelling in coding and intricate task analysis. Furthermore, it boasts a generous context length of 128 K tokens, compatibility with tool-calling APIs, and support for industry-standard inference engines, making it a versatile option for various applications. The innovative design and features of Kimi K2 position it as a significant advancement in the field of artificial intelligence language processing.

On September 17, 2024, Mistral AI revealed a series of significant updates designed to improve both the accessibility and efficiency of their AI products. Among these updates was the introduction of a complimentary tier on "La Plateforme," their serverless platform that allows for the tuning and deployment of Mistral models as API endpoints, which gives developers a chance to innovate and prototype at zero cost. In addition, Mistral AI announced price reductions across their complete model range, highlighted by a remarkable 50% decrease for Mistral Nemo and an 80% cut for Mistral Small and Codestral, thereby making advanced AI solutions more affordable for a wider audience. The company also launched Mistral Small v24.09, a model with 22 billion parameters that strikes a favorable balance between performance and efficiency, making it ideal for various applications such as translation, summarization, and sentiment analysis. Moreover, they released Pixtral 12B, a vision-capable model equipped with image understanding features, for free on "Le Chat," allowing users to analyze and caption images while maintaining strong text-based performance. This suite of updates reflects Mistral AI's commitment to democratizing access to powerful AI technologies for developers everywhere.

SWE-1.6 is a cutting-edge AI model focused on engineering, created by Cognition and embedded within the Windsurf environment, with the goal of enhancing both the raw intelligence and what Cognition refers to as “model UX,” which encompasses the overall user interaction experience with the AI. This latest version marks a significant upgrade in the SWE model series, boasting a performance increase of over 10% on benchmarks like SWE-Bench Pro when compared to its predecessor, SWE-1.5, all while retaining similar foundational capabilities. Developed from the ground up, it aims to elevate both reasoning quality and user satisfaction, effectively tackling challenges identified in previous iterations, such as overanalyzing straightforward questions, excessive steps in problem-solving, repetitive reasoning loops, and an overreliance on terminal commands rather than utilizing specialized tools. The enhancements introduced in SWE-1.6 include improved behaviors such as a greater frequency of simultaneous tool usage, quicker context retrieval, and a diminished necessity for user input, leading to more fluid and productive workflows. In addition, these refinements contribute to a more intuitive interaction for users, ensuring that tasks can be completed with greater ease and efficiency than ever before.

GPT-4.1 mini is a streamlined version of GPT-4.1, offering the same core capabilities in coding, instruction adherence, and long-context comprehension, but with faster performance and lower costs. Ideal for developers seeking to integrate AI into real-time applications, GPT-4.1 mini maintains a 1 million token context window and is well-suited for tasks that demand low-latency responses. It is a cost-effective option for businesses that need powerful AI capabilities without the high overhead associated with larger models.

OLMo 2 represents a collection of completely open language models created by the Allen Institute for AI (AI2), aimed at giving researchers and developers clear access to training datasets, open-source code, reproducible training methodologies, and thorough assessments. These models are trained on an impressive volume of up to 5 trillion tokens and compete effectively with top open-weight models like Llama 3.1, particularly in English academic evaluations. A key focus of OLMo 2 is on ensuring training stability, employing strategies to mitigate loss spikes during extended training periods, and applying staged training interventions in the later stages of pretraining to mitigate weaknesses in capabilities. Additionally, the models leverage cutting-edge post-training techniques derived from AI2's Tülu 3, leading to the development of OLMo 2-Instruct models. To facilitate ongoing enhancements throughout the development process, an actionable evaluation framework known as the Open Language Modeling Evaluation System (OLMES) was created, which includes 20 benchmarks that evaluate essential capabilities. This comprehensive approach not only fosters transparency but also encourages continuous improvement in language model performance.

Mistral AI has introduced the Mistral Large 2, a sophisticated AI model crafted to excel in various domains such as code generation, multilingual understanding, and intricate reasoning tasks. With an impressive 128k context window, this model accommodates a wide array of languages, including English, French, Spanish, and Arabic, while also supporting an extensive list of over 80 programming languages. Designed for high-throughput single-node inference, Mistral Large 2 is perfectly suited for applications requiring large context handling. Its superior performance on benchmarks like MMLU, coupled with improved capabilities in code generation and reasoning, guarantees both accuracy and efficiency in results. Additionally, the model features enhanced function calling and retrieval mechanisms, which are particularly beneficial for complex business applications. This makes Mistral Large 2 not only versatile but also a powerful tool for developers and businesses looking to leverage advanced AI capabilities.