Alternatives to LTM-2-mini

MiniMax M3 is a frontier open-weight AI model built for coding, agentic work, multimodal understanding, and ultra-long-context tasks. The model supports up to a 1 million token context window, allowing it to work across large codebases, long documents, logs, project histories, and complex task environments. MiniMax M3 introduces MiniMax Sparse Attention, a sparse attention architecture designed to make long-context processing more efficient. The model is natively multimodal, with training that supports deeper semantic fusion across text, image, and video inputs. It is designed to support software engineering tasks, repository analysis, terminal-style work, browser-style retrieval, tool use, and autonomous workflows. MiniMax M3 has a mixture-of-experts architecture with hundreds of billions of total parameters and a smaller activated parameter count for more efficient inference. Developers can use it for AI coding assistants, workflow automation, research agents, document analysis, visual reasoning, and enterprise AI systems. Its long-context capability makes it especially useful when tasks require many files, references, instructions, or interaction histories to stay available at once. MiniMax M3 helps teams build more capable AI agents that can understand larger problems, work across multiple modalities, and execute complex tasks with stronger context awareness.

Kimi K3 is a large-scale AI model from Moonshot AI designed for advanced reasoning, software engineering, visual understanding, agentic workflows, and knowledge work. The model is built with 2.8 trillion parameters and uses Kimi Delta Attention, a hybrid linear attention design created to support long-context intelligence. It also includes Attention Residuals and a native 1 million token context window, giving developers room to work with large files, repositories, documentation sets, transcripts, and enterprise knowledge bases. Kimi K3 always runs with thinking mode enabled and currently supports maximum reasoning effort by default. Developers can access the model through Moonshot’s OpenAI-compatible API using Python, cURL, and the OpenAI SDK. The API supports standard chat completions, streaming output, structured JSON Schema responses, partial continuation from a prefix, custom tool calling, required tool choice, and dynamic tool loading. Kimi K3 also supports vision inputs, including local images encoded as base64 and video files uploaded through the file API. Automatic context caching helps repeated long-prefix workflows become more efficient without requiring manual cache IDs or extra cache parameters. By combining long context, visual understanding, tool use, structured output, and advanced reasoning, Kimi K3 is built for developers creating sophisticated AI agents, coding systems, research tools, and enterprise applications.

OpenAI’s GPT-5 mini is a cost-efficient, faster version of the flagship GPT-5 model, designed to handle well-defined tasks and precise inputs with high reasoning capabilities. Supporting text and image inputs, GPT-5 mini can process and generate large amounts of content thanks to its extensive 400,000-token context window and a maximum output of 128,000 tokens. This model is optimized for speed, making it ideal for developers and businesses needing quick turnaround times on natural language processing tasks while maintaining accuracy. The pricing model offers significant savings, charging $0.25 per million input tokens and $2 per million output tokens, compared to the higher costs of the full GPT-5. It supports many advanced API features such as streaming responses, function calling, and fine-tuning, while excluding audio input and image generation capabilities. GPT-5 mini is compatible with a broad range of API endpoints including chat completions, real-time responses, and embeddings, making it highly flexible. Rate limits vary by usage tier, supporting from hundreds to tens of thousands of requests per minute, ensuring reliability for different scale needs. This model strikes a balance between performance and cost, suitable for applications requiring fast, high-quality AI interaction without extensive resource use.

DeepSeek-V4-Pro is an advanced Mixture-of-Experts language model built for high-performance reasoning, coding, and large-scale AI applications. With 1.6 trillion total parameters and 49 billion activated parameters, it delivers strong capabilities while maintaining computational efficiency. The model supports a massive context window of up to one million tokens, making it ideal for handling long documents and complex workflows. Its hybrid attention architecture improves efficiency by reducing computational overhead while maintaining accuracy. Trained on more than 32 trillion tokens, DeepSeek-V4-Pro demonstrates strong performance across knowledge, reasoning, and coding benchmarks. It includes advanced training techniques such as improved optimization and enhanced signal propagation for better stability. The model offers multiple reasoning modes, allowing users to choose between faster responses or deeper analytical thinking. It is designed to support agentic workflows and complex multi-step problem solving. As an open-source model, it provides flexibility for developers and organizations to customize and deploy at scale. Overall, DeepSeek-V4-Pro delivers a balance of performance, efficiency, and scalability for demanding AI applications.

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.

A compact model that excels in textual understanding and multimodal reasoning capabilities. The GPT-4o mini is designed to handle a wide array of tasks efficiently, thanks to its low cost and minimal latency, making it ideal for applications that require chaining or parallelizing multiple model calls, such as invoking several APIs simultaneously, processing extensive context like entire codebases or conversation histories, and providing swift, real-time text interactions for customer support chatbots. Currently, the API for GPT-4o mini accommodates both text and visual inputs, with plans to introduce support for text, images, videos, and audio in future updates. This model boasts an impressive context window of 128K tokens and can generate up to 16K output tokens per request, while its knowledge base is current as of October 2023. Additionally, the enhanced tokenizer shared with GPT-4o has made it more efficient in processing non-English text, further broadening its usability for diverse applications. As a result, GPT-4o mini stands out as a versatile tool for developers and businesses alike.

Phi-4-mini-flash-reasoning is a 3.8 billion-parameter model that is part of Microsoft's Phi series, specifically designed for edge, mobile, and other environments with constrained resources where processing power, memory, and speed are limited. This innovative model features the SambaY hybrid decoder architecture, integrating Gated Memory Units (GMUs) with Mamba state-space and sliding-window attention layers, achieving up to ten times the throughput and a latency reduction of 2 to 3 times compared to its earlier versions without compromising on its ability to perform complex mathematical and logical reasoning. With a support for a context length of 64K tokens and being fine-tuned on high-quality synthetic datasets, it is particularly adept at handling long-context retrieval, reasoning tasks, and real-time inference, all manageable on a single GPU. Available through platforms such as Azure AI Foundry, NVIDIA API Catalog, and Hugging Face, Phi-4-mini-flash-reasoning empowers developers to create applications that are not only fast but also scalable and capable of intensive logical processing. This accessibility allows a broader range of developers to leverage its capabilities for innovative solutions.

Introducing the next iteration of our open-source large language model, this version features model weights along with initial code for the pretrained and fine-tuned Llama language models, which span from 7 billion to 70 billion parameters. The Llama 2 pretrained models have been developed using an impressive 2 trillion tokens and offer double the context length compared to their predecessor, Llama 1. Furthermore, the fine-tuned models have been enhanced through the analysis of over 1 million human annotations. Llama 2 demonstrates superior performance against various other open-source language models across multiple external benchmarks, excelling in areas such as reasoning, coding capabilities, proficiency, and knowledge assessments. For its training, Llama 2 utilized publicly accessible online data sources, while the fine-tuned variant, Llama-2-chat, incorporates publicly available instruction datasets along with the aforementioned extensive human annotations. Our initiative enjoys strong support from a diverse array of global stakeholders who are enthusiastic about our open approach to AI, including companies that have provided valuable early feedback and are eager to collaborate using Llama 2. The excitement surrounding Llama 2 signifies a pivotal shift in how AI can be developed and utilized collectively.

Llama 4 Scout is an advanced multimodal AI model with 17 billion active parameters, offering industry-leading performance with a 10 million token context length. This enables it to handle complex tasks like multi-document summarization and detailed code reasoning with impressive accuracy. Scout surpasses previous Llama models in both text and image understanding, making it an excellent choice for applications that require a combination of language processing and image analysis. Its powerful capabilities in long-context tasks and image-grounding applications set it apart from other models in its class, providing superior results for a wide range of industries.

Phi-4-mini-reasoning is a transformer-based language model with 3.8 billion parameters, specifically designed to excel in mathematical reasoning and methodical problem-solving within environments that have limited computational capacity or latency constraints. Its optimization stems from fine-tuning with synthetic data produced by the DeepSeek-R1 model, striking a balance between efficiency and sophisticated reasoning capabilities. With training that encompasses over one million varied math problems, ranging in complexity from middle school to Ph.D. level, Phi-4-mini-reasoning demonstrates superior performance to its base model in generating lengthy sentences across multiple assessments and outshines larger counterparts such as OpenThinker-7B, Llama-3.2-3B-instruct, and DeepSeek-R1. Equipped with a 128K-token context window, it also facilitates function calling, which allows for seamless integration with various external tools and APIs. Moreover, Phi-4-mini-reasoning can be quantized through the Microsoft Olive or Apple MLX Framework, enabling its deployment on a variety of edge devices, including IoT gadgets, laptops, and smartphones. Its design not only enhances user accessibility but also expands the potential for innovative applications in mathematical fields.

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.

The MiniMax‑M1 model, introduced by MiniMax AI and licensed under Apache 2.0, represents a significant advancement in hybrid-attention reasoning architecture. With an extraordinary capacity for handling a 1 million-token context window and generating outputs of up to 80,000 tokens, it facilitates in-depth analysis of lengthy texts. Utilizing a cutting-edge CISPO algorithm, MiniMax‑M1 was trained through extensive reinforcement learning, achieving completion on 512 H800 GPUs in approximately three weeks. This model sets a new benchmark in performance across various domains, including mathematics, programming, software development, tool utilization, and understanding of long contexts, either matching or surpassing the capabilities of leading models in the field. Additionally, users can choose between two distinct variants of the model, each with a thinking budget of either 40K or 80K, and access the model's weights and deployment instructions on platforms like GitHub and Hugging Face. Such features make MiniMax‑M1 a versatile tool for developers and researchers alike.

Mistral Small 3.1 represents a cutting-edge, multimodal, and multilingual AI model that has been released under the Apache 2.0 license. This upgraded version builds on Mistral Small 3, featuring enhanced text capabilities and superior multimodal comprehension, while also accommodating an extended context window of up to 128,000 tokens. It demonstrates superior performance compared to similar models such as Gemma 3 and GPT-4o Mini, achieving impressive inference speeds of 150 tokens per second. Tailored for adaptability, Mistral Small 3.1 shines in a variety of applications, including instruction following, conversational support, image analysis, and function execution, making it ideal for both business and consumer AI needs. The model's streamlined architecture enables it to operate efficiently on hardware such as a single RTX 4090 or a Mac equipped with 32GB of RAM, thus supporting on-device implementations. Users can download it from Hugging Face and access it through Mistral AI's developer playground, while it is also integrated into platforms like Gemini Enterprise Agent Platform, with additional accessibility on NVIDIA NIM and more. This flexibility ensures that developers can leverage its capabilities across diverse environments and applications.

Designed for power users, OpenAI o4-mini-high is the go-to model when you need the best balance of performance and cost-efficiency. With its improved reasoning abilities, o4-mini-high excels in high-volume tasks that require advanced data analysis, algorithm optimization, and multi-step reasoning. It's ideal for businesses or developers who need to scale their AI solutions without sacrificing speed or accuracy.

The o4-mini model, a more compact and efficient iteration of the o3 model, was developed to enhance reasoning capabilities and streamline performance. It excels in tasks requiring complex problem-solving, making it an ideal solution for users demanding more powerful AI. By refining its design, OpenAI has made significant strides in creating a model that balances efficiency with advanced capabilities. With this release, the o4-mini is poised to meet the growing need for smarter AI tools while maintaining the robust functionality of its predecessor. It plays a critical role in OpenAI’s ongoing efforts to push the boundaries of artificial intelligence ahead of the GPT-5 launch.

The TinyLlama initiative seeks to pretrain a Llama model with 1.1 billion parameters using a dataset of 3 trillion tokens. With the right optimizations, this ambitious task can be completed in a mere 90 days, utilizing 16 A100-40G GPUs. We have maintained the same architecture and tokenizer as Llama 2, ensuring that TinyLlama is compatible with various open-source projects that are based on Llama. Additionally, the model's compact design, consisting of just 1.1 billion parameters, makes it suitable for numerous applications that require limited computational resources and memory. This versatility enables developers to integrate TinyLlama seamlessly into their existing frameworks and workflows.

SubQ 1.1 Small is the second iteration of Subquadratic’s long-context AI model, built to help enterprises solve problems that require reasoning across entire artifacts rather than isolated chunks. The model is designed for use cases involving large code repositories, document libraries, legal agreements, financial reports, contracts, and other complex information sets. Its Subquadratic Sparse Attention architecture reduces the compute burden of traditional dense attention, making it more practical to process multi-million-token contexts. SubQ 1.1 Small achieves near-perfect performance on needle-in-a-haystack retrieval tests up to 12M tokens, despite being trained primarily at 1M tokens. It also performs strongly on RULER, GPQA Diamond, LiveCodeBench, and AutomationBench Finance, showing a balance between long-context retrieval and general reasoning ability. At 1M tokens, the model uses 64.5x less compute than dense attention and runs 56x faster than FlashAttention-2 on a single attention layer. This efficiency makes long-context training and inference more scalable for enterprise AI applications. SubQ 1.1 Small is especially valuable for teams that need to analyze relationships across full documents, trace logic across codebases, or connect information across extensive collections. The model is intended to help organizations reduce dependence on complex retrieval workarounds and reason more directly over large-scale data.

The o3-mini by OpenAI is a streamlined iteration of the sophisticated o3 AI model, delivering robust reasoning skills in a more compact and user-friendly format. It specializes in simplifying intricate instructions into digestible steps, making it particularly adept at coding, competitive programming, and tackling mathematical and scientific challenges. This smaller model maintains the same level of accuracy and logical reasoning as the larger version, while operating with lower computational demands, which is particularly advantageous in environments with limited resources. Furthermore, o3-mini incorporates inherent deliberative alignment, promoting safe, ethical, and context-sensitive decision-making. Its versatility makes it an invaluable resource for developers, researchers, and enterprises striving for an optimal mix of performance and efficiency in their projects. The combination of these features positions o3-mini as a significant tool in the evolving landscape of AI-driven solutions.

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.

The Grok-3 Mini, developed by xAI, serves as a nimble and perceptive AI assistant specifically designed for individuals seeking prompt yet comprehensive responses to their inquiries. Retaining the core attributes of the Grok series, this compact variant offers a lighthearted yet insightful viewpoint on various human experiences while prioritizing efficiency. It caters to those who are constantly on the go or have limited access to resources, ensuring that the same level of inquisitiveness and support is delivered in a smaller package. Additionally, Grok-3 Mini excels at addressing a wide array of questions, offering concise insights without sacrificing depth or accuracy, which makes it an excellent resource for navigating the demands of contemporary life. Ultimately, it embodies a blend of practicality and intelligence that meets the needs of modern users.

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.

Amazon Nova Micro is an advanced text-only AI model optimized for rapid language processing at a very low cost. With capabilities in reasoning, translation, and code completion, it offers over 200 tokens per second in response generation, making it suitable for fast-paced, real-time applications. Nova Micro supports fine-tuning with text inputs, and its efficiency in understanding and generating text makes it a cost-effective solution for AI-driven applications requiring high performance and quick outputs.

MiniMax is an AI platform that provides multimodal foundation models, developer tools, and intelligent agent solutions designed to support coding, automation, content generation, and enterprise AI applications. The company’s flagship model, MiniMax M3, delivers advanced coding performance, long-context processing capabilities of up to one million tokens, and native multimodal functionality that enables seamless understanding and generation of text, audio, images, video, and music. MiniMax Code serves as an AI-powered coding environment that allows users to build agent teams, automate repetitive development tasks, create custom skills, and manage workflows through a unified conversational interface. In addition to coding solutions, the platform offers video generation through Hailuo AI, speech and music generation models, conversational AI products, and developer APIs for integrating AI into custom applications. The platform is designed to support both individual users and enterprise teams seeking scalable AI tools for software development, business automation, creative production, and research. By combining frontier AI models with practical productivity applications, MiniMax enables organizations to streamline operations, enhance innovation, and build intelligent systems more efficiently.

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.

Code Llama is an advanced language model designed to generate code through text prompts, distinguishing itself as a leading tool among publicly accessible models for coding tasks. This innovative model not only streamlines workflows for existing developers but also aids beginners in overcoming challenges associated with learning to code. Its versatility positions Code Llama as both a valuable productivity enhancer and an educational resource, assisting programmers in creating more robust and well-documented software solutions. Additionally, users can generate both code and natural language explanations by providing either type of prompt, making it an adaptable tool for various programming needs. Available for free for both research and commercial applications, Code Llama is built upon Llama 2 architecture and comes in three distinct versions: the foundational Code Llama model, Code Llama - Python which is tailored specifically for Python programming, and Code Llama - Instruct, optimized for comprehending and executing natural language directives effectively.

The o3-mini-high model developed by OpenAI enhances artificial intelligence reasoning capabilities by improving deep problem-solving skills in areas such as programming, mathematics, and intricate tasks. This model incorporates adaptive thinking time and allows users to select from various reasoning modes—low, medium, and high—to tailor performance to the difficulty of the task at hand. Impressively, it surpasses the o1 series by an impressive 200 Elo points on Codeforces, providing exceptional efficiency at a reduced cost while ensuring both speed and precision in its operations. As a notable member of the o3 family, this model not only expands the frontiers of AI problem-solving but also remains user-friendly, offering a complimentary tier alongside increased limits for Plus subscribers, thereby making advanced AI more widely accessible. Its innovative design positions it as a significant tool for users looking to tackle challenging problems with enhanced support and adaptability.

OpenAI’s GPT-5 nano is the most cost-effective and rapid variant of the GPT-5 series, tailored for tasks like summarization, classification, and other well-defined language problems. Supporting both text and image inputs, GPT-5 nano can handle extensive context lengths of up to 400,000 tokens and generate detailed outputs of up to 128,000 tokens. Its emphasis on speed makes it ideal for applications that require quick, reliable AI responses without the resource demands of larger models. With highly affordable pricing — just $0.05 per million input tokens and $0.40 per million output tokens — GPT-5 nano is accessible to a wide range of developers and businesses. The model supports key API functionalities including streaming responses, function calling, structured output, and fine-tuning capabilities. While it does not support web search or audio input, it efficiently handles code interpretation, image generation, and file search tasks. Rate limits scale with usage tiers to ensure reliable access across small to enterprise deployments. GPT-5 nano offers an excellent balance of speed, affordability, and capability for lightweight AI applications.

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.

We have incorporated Llama 3 into Meta AI, our intelligent assistant that enhances how individuals accomplish tasks, innovate, and engage with Meta AI. By utilizing Meta AI for coding and problem-solving, you can experience Llama 3's capabilities first-hand. Whether you are creating agents or other AI-driven applications, Llama 3, available in both 8B and 70B versions, will provide the necessary capabilities and flexibility to bring your ideas to fruition. With the launch of Llama 3, we have also revised our Responsible Use Guide (RUG) to offer extensive guidance on the ethical development of LLMs. Our system-focused strategy encompasses enhancements to our trust and safety mechanisms, including Llama Guard 2, which is designed to align with the newly introduced taxonomy from MLCommons, broadening its scope to cover a wider array of safety categories, alongside code shield and Cybersec Eval 2. Additionally, these advancements aim to ensure a safer and more responsible use of AI technologies in various applications.

MiniMax M2 is an open-source foundational model tailored for agent-driven applications and coding tasks, achieving an innovative equilibrium of efficiency, velocity, and affordability. It shines in comprehensive development environments, adeptly managing programming tasks, invoking tools, and executing intricate, multi-step processes, complete with features like Python integration, while offering impressive inference speeds of approximately 100 tokens per second and competitive API pricing at around 8% of similar proprietary models. The model includes a "Lightning Mode" designed for rapid, streamlined agent operations, alongside a "Pro Mode" aimed at thorough full-stack development, report creation, and the orchestration of web-based tools; its weights are entirely open source, allowing for local deployment via vLLM or SGLang. MiniMax M2 stands out as a model ready for production use, empowering agents to autonomously perform tasks such as data analysis, software development, tool orchestration, and implementing large-scale, multi-step logic across real organizational contexts. With its advanced capabilities, this model is poised to revolutionize the way developers approach complex programming challenges.

StarCoder and StarCoderBase represent advanced Large Language Models specifically designed for code, developed using openly licensed data from GitHub, which encompasses over 80 programming languages, Git commits, GitHub issues, and Jupyter notebooks. In a manner akin to LLaMA, we constructed a model with approximately 15 billion parameters trained on a staggering 1 trillion tokens. Furthermore, we tailored the StarCoderBase model with 35 billion Python tokens, leading to the creation of what we now refer to as StarCoder. Our evaluations indicated that StarCoderBase surpasses other existing open Code LLMs when tested against popular programming benchmarks and performs on par with or even exceeds proprietary models like code-cushman-001 from OpenAI, the original Codex model that fueled early iterations of GitHub Copilot. With an impressive context length exceeding 8,000 tokens, the StarCoder models possess the capability to handle more information than any other open LLM, thus paving the way for a variety of innovative applications. This versatility is highlighted by our ability to prompt the StarCoder models through a sequence of dialogues, effectively transforming them into dynamic technical assistants that can provide support in diverse programming tasks.

DeepSeek-V4 is an advanced open-source large language model engineered for efficient long-context processing and high-level reasoning tasks. Supporting a massive one million token context window, it enables developers to build applications that handle extensive data and complex workflows without fragmentation. The model is available in two versions: V4-Pro for maximum reasoning power and V4-Flash for faster, cost-efficient performance. DeepSeek-V4-Pro delivers top-tier results in coding, mathematics, and knowledge benchmarks, rivaling leading proprietary models. Its architecture incorporates innovative attention techniques that significantly improve efficiency while maintaining strong performance. The model is optimized for agent-based workflows, allowing seamless integration with tools and automation systems. It also supports dual reasoning modes, enabling users to switch between quick responses and deeper analytical outputs. DeepSeek-V4 is fully open-source, providing flexibility for customization and deployment across various environments. Overall, it offers a powerful and scalable solution for modern AI development.

Inkling-Small is an efficient Mixture-of-Experts transformer model built to provide performance comparable to Inkling while using a much smaller active parameter footprint. The model has 276 billion total parameters and 12 billion active parameters, making it designed for strong capability with more efficient compute usage. Inkling-Small was trained on NVIDIA GB300 NVL72 systems and supports native reasoning across text, images, and audio. It offers context windows of up to one million tokens, making it suitable for long documents, large codebases, multimodal context, and extended agent workflows. Users can set reasoning effort from minimal to extra high to control the balance between speed, cost, compute, and task complexity. The model benefits from improved pre-training data, post-training with on-policy distillation from Inkling, and extended agentic coding reinforcement learning. These improvements helped Inkling-Small surpass its larger counterpart on reasoning and coding benchmarks. Its encoder-free multimodal architecture processes audio as dMel spectrograms and images as 40-by-40-pixel patches alongside text tokens. By combining efficient MoE scaling, long-context reasoning, multimodal input, coding strength, and adjustable thinking effort, Inkling-Small is built for practical high-performance AI deployment.

We are excited to present MPT-7B, the newest addition to the MosaicML Foundation Series. This transformer model has been meticulously trained from the ground up using 1 trillion tokens of diverse text and code. It is open-source and ready for commercial applications, delivering performance on par with LLaMA-7B. The training process took 9.5 days on the MosaicML platform, requiring no human input and incurring an approximate cost of $200,000. With MPT-7B, you can now train, fine-tune, and launch your own customized MPT models, whether you choose to begin with one of our provided checkpoints or start anew. To provide additional options, we are also introducing three fine-tuned variants alongside the base MPT-7B: MPT-7B-Instruct, MPT-7B-Chat, and MPT-7B-StoryWriter-65k+, the latter boasting an impressive context length of 65,000 tokens, allowing for extensive content generation. These advancements open up new possibilities for developers and researchers looking to leverage the power of transformer models in their projects.

GPT-5.4 mini is an advanced AI model designed to provide a balance between high performance, speed, and cost efficiency. It is built to handle a wide range of tasks, including coding, reasoning, tool usage, and multimodal understanding. Compared to earlier versions, GPT-5.4 mini delivers significantly improved performance while operating at faster speeds. The model is particularly effective in environments where low latency is essential, such as real-time coding assistants and interactive applications. It supports capabilities like function calling, tool integration, and image-based reasoning, making it highly versatile. GPT-5.4 mini is also well-suited for subagent architectures, where it can efficiently process smaller tasks within larger AI systems. Developers can use it to automate workflows, analyze data, and build responsive AI-driven applications. Its strong performance across benchmarks shows that it approaches the capabilities of larger models in many scenarios. At the same time, it maintains a lower cost, making it ideal for high-volume usage. Overall, GPT-5.4 mini provides a powerful and scalable solution for modern AI development.

DeepSeek-V4-Flash is an optimized Mixture-of-Experts language model built for efficient large-scale AI workloads and fast inference. With 284 billion total parameters and 13 billion activated parameters, it delivers strong performance while maintaining lower computational demands compared to larger models. The model supports a massive context length of up to one million tokens, making it suitable for handling long-form content and multi-step workflows. Its hybrid attention mechanism improves efficiency by minimizing resource consumption while preserving accuracy. Trained on a dataset exceeding 32 trillion tokens, DeepSeek-V4-Flash performs well across reasoning, coding, and knowledge benchmarks. It offers flexible reasoning modes, enabling users to switch between quick responses and more detailed analytical outputs. The architecture is designed to support agentic workflows and scalable deployment environments. As an open-source model, it provides flexibility for customization and integration. Overall, DeepSeek-V4-Flash is a cost-effective and high-performance solution for modern AI applications.

DeepSeek-V2 is a cutting-edge Mixture-of-Experts (MoE) language model developed by DeepSeek-AI, noted for its cost-effective training and high-efficiency inference features. It boasts an impressive total of 236 billion parameters, with only 21 billion active for each token, and is capable of handling a context length of up to 128K tokens. The model utilizes advanced architectures such as Multi-head Latent Attention (MLA) to optimize inference by minimizing the Key-Value (KV) cache and DeepSeekMoE to enable economical training through sparse computations. Compared to its predecessor, DeepSeek 67B, this model shows remarkable improvements, achieving a 42.5% reduction in training expenses, a 93.3% decrease in KV cache size, and a 5.76-fold increase in generation throughput. Trained on an extensive corpus of 8.1 trillion tokens, DeepSeek-V2 demonstrates exceptional capabilities in language comprehension, programming, and reasoning tasks, positioning it as one of the leading open-source models available today. Its innovative approach not only elevates its performance but also sets new benchmarks within the field of artificial intelligence.

CodeQwen serves as the coding counterpart to Qwen, which is a series of large language models created by the Qwen team at Alibaba Cloud. Built on a transformer architecture that functions solely as a decoder, this model has undergone extensive pre-training using a vast dataset of code. It showcases robust code generation abilities and demonstrates impressive results across various benchmarking tests. With the capacity to comprehend and generate long contexts of up to 64,000 tokens, CodeQwen accommodates 92 programming languages and excels in tasks such as text-to-SQL queries and debugging. Engaging with CodeQwen is straightforward—you can initiate a conversation with just a few lines of code utilizing transformers. The foundation of this interaction relies on constructing the tokenizer and model using pre-existing methods, employing the generate function to facilitate dialogue guided by the chat template provided by the tokenizer. In alignment with our established practices, we implement the ChatML template tailored for chat models. This model adeptly completes code snippets based on the prompts it receives, delivering responses without the need for any further formatting adjustments, thereby enhancing the user experience. The seamless integration of these elements underscores the efficiency and versatility of CodeQwen in handling diverse coding tasks.

Seed2.0 Mini represents the most compact version of ByteDance's Seed2.0 line of versatile multimodal agent models, crafted for efficient high-throughput inference and dense deployment, while still embodying the essential strengths found in its larger counterparts regarding multimodal understanding and instruction adherence. This Mini variant, alongside Pro and Lite siblings, is particularly fine-tuned for handling high-concurrency and batch generation tasks, proving itself ideal for scenarios where the ability to process numerous requests simultaneously is as crucial as its overall capability. In line with other models in the Seed2.0 family, it showcases notable improvements in visual reasoning and motion perception, excels at extracting structured information from intricate inputs such as text and images, and effectively carries out multi-step instructions. However, in exchange for enhanced inference speed and cost efficiency, it sacrifices some degree of raw reasoning power and output quality, ensuring that it remains a practical option for various applications. As a result, Seed2.0 Mini strikes a balance between performance and efficiency, appealing to developers seeking to optimize their systems for scalable solutions.

The Nemotron-3 Super is an innovative member of NVIDIA's Nemotron 3 series of open models, specifically crafted to facilitate sophisticated agentic AI systems that can effectively reason, plan, and carry out multi-step workflows in intricate environments. This model features a unique hybrid Mamba-Transformer Mixture-of-Experts architecture that merges the streamlined efficiency of Mamba layers with the contextual depth provided by transformer attention mechanisms, which allows it to adeptly manage extended sequences and intricate reasoning tasks with impressive accuracy and throughput. By activating only a portion of its parameters for each token, this architecture significantly enhances computational efficiency while preserving robust reasoning capabilities, making it ideal for scalable inference under heavy workloads. The Nemotron-3 Super comprises approximately 120 billion parameters, with around 12 billion being active during inference, which substantially boosts its ability to handle multi-step reasoning and collaborative interactions among agents within extensive contexts. Such advancements make it a powerful tool for tackling diverse challenges in AI applications.

Falcon-7B is a causal decoder-only model comprising 7 billion parameters, developed by TII and trained on an extensive dataset of 1,500 billion tokens from RefinedWeb, supplemented with specially selected corpora, and it is licensed under Apache 2.0. What are the advantages of utilizing Falcon-7B? This model surpasses similar open-source alternatives, such as MPT-7B, StableLM, and RedPajama, due to its training on a remarkably large dataset of 1,500 billion tokens from RefinedWeb, which is further enhanced with carefully curated content, as evidenced by its standing on the OpenLLM Leaderboard. Additionally, it boasts an architecture that is finely tuned for efficient inference, incorporating technologies like FlashAttention and multiquery mechanisms. Moreover, the permissive nature of the Apache 2.0 license means users can engage in commercial applications without incurring royalties or facing significant limitations. This combination of performance and flexibility makes Falcon-7B a strong choice for developers seeking advanced modeling capabilities.

Xiaomi MiMo-V2.5-Pro is a next-generation open-source AI model designed for advanced reasoning, coding, and long-horizon task execution. It uses a Mixture-of-Experts architecture with over one trillion parameters and a large active parameter set for efficient performance. The model supports an extended context window of up to one million tokens, allowing it to handle complex, multi-step workflows. It is built to perform autonomous tasks, including software development, system design, and engineering optimization. Benchmark results show strong performance across coding, reasoning, and agent-based evaluation tests. MiMo-V2.5-Pro incorporates hybrid attention mechanisms to improve efficiency while maintaining accuracy across long contexts. It is optimized for token efficiency, reducing the computational cost of running complex tasks. The model can integrate with development tools and frameworks to support real-world applications. It is designed to complete tasks that would typically require significant human effort over extended periods. Xiaomi has made the model open source, enabling developers to access and customize it. By combining performance, scalability, and efficiency, MiMo-V2.5-Pro pushes the boundaries of modern AI capabilities.

Solar Mini is an advanced pre-trained large language model that matches the performance of GPT-3.5 while providing responses 2.5 times faster, all while maintaining a parameter count of under 30 billion. In December 2023, it secured the top position on the Hugging Face Open LLM Leaderboard by integrating a 32-layer Llama 2 framework, which was initialized with superior Mistral 7B weights, coupled with a novel method known as "depth up-scaling" (DUS) that enhances the model's depth efficiently without the need for intricate modules. Following the DUS implementation, the model undergoes further pretraining to restore and boost its performance, and it also includes instruction tuning in a question-and-answer format, particularly tailored for Korean, which sharpens its responsiveness to user prompts, while alignment tuning ensures its outputs align with human or sophisticated AI preferences. Solar Mini consistently surpasses rivals like Llama 2, Mistral 7B, Ko-Alpaca, and KULLM across a range of benchmarks, demonstrating that a smaller model can still deliver exceptional performance. This showcases the potential of innovative architectural strategies in the development of highly efficient AI models.

MiniMax M2.7 is a powerful AI model built to drive real-world productivity across coding, search, and office-based workflows. It is trained using reinforcement learning across a wide range of real-world environments, enabling it to execute complex, multi-step tasks with precision and efficiency. The model demonstrates strong problem-solving capabilities by breaking down challenges into structured steps before generating solutions across multiple programming languages. It delivers high-speed performance with rapid token output, ensuring faster completion of demanding tasks. With optimized reasoning, it reduces token usage and execution time, making it more efficient than previous models. M2.7 also achieves state-of-the-art results in software engineering benchmarks, significantly improving response times for technical issues. Its advanced agentic capabilities allow it to work seamlessly with tools and support complex workflows with high skill accuracy. The model is designed to handle professional tasks, including multi-turn interactions and high-quality document editing. It also provides strong support for office productivity, enabling efficient handling of structured data and business tasks. With competitive pricing, it delivers high performance while remaining cost-effective. Overall, it combines speed, intelligence, and versatility to meet the needs of modern professionals and teams.

Qwen3.5-35B-A3B is a member of the Qwen3.5 "Medium" model series, meticulously crafted as an effective multimodal foundation model that strikes a balance between robust reasoning capabilities and practical application needs. Utilizing a Mixture-of-Experts (MoE) architecture, it boasts a total of 35 billion parameters, yet activates only around 3 billion for each token, enabling it to achieve performance levels similar to much larger models while significantly cutting down on computational expenses. The model employs a hybrid attention mechanism that merges linear attention with traditional attention layers, which enhances its ability to handle extensive context and boosts scalability for intricate tasks. As an inherently vision-language model, it processes both textual and visual data, catering to a variety of applications, including multimodal reasoning, programming, and automated workflows. Furthermore, it is engineered to operate as a versatile "AI agent," proficient in planning, utilizing tools, and systematically solving problems, extending its functionality beyond mere conversational interactions. This capability positions it as a valuable asset across diverse domains, where advanced AI-driven solutions are increasingly required.