OpenRouter Integrations

MacWhisper is a Mac transcription and dictation app that helps users transcribe audio, video, meetings, podcasts, lectures, interviews, subtitles, voice memos, and private files. The app supports drag-and-drop transcription for common media formats and can record meetings from Zoom, Teams, Webex, Skype, Chime, Discord, and other online meeting tools. MacWhisper can also capture and transcribe audio from any app on a Mac, making it useful for videos, calls, recordings, and media workflows. The platform is built with privacy in mind, offering local AI models and offline processing for sensitive content. Users can generate accurate transcripts, recognize speakers, remove filler words, translate text, search transcripts, edit content, and export files in formats such as subtitles, text, Markdown, PDF, HTML, and DOCX. Batch transcription helps professionals process multiple files at once. MacWhisper Pro adds AI services, custom prompts, cloud and local model options, app-specific dictation prompts, automatic meeting detection, watched folders, workflow uploads, and CLI control. The app can connect to AI providers such as OpenAI, Anthropic, xAI, Google Gemini, DeepSeek, Azure, OpenRouter, Ollama, LM Studio, Deepgram, ElevenLabs, and others. By combining transcription, meeting recording, dictation, privacy-focused local processing, AI summaries, exports, integrations, and workflow automation, MacWhisper helps users turn spoken content into useful text.

RA.Aid is an open-source AI assistant that streamlines research, planning, and execution to accelerate software development workflows. Utilizing LangGraph's agent-based task management structure, RA.Aid functions through a three-tier architecture. It is compatible with various AI providers, such as Anthropic's Claude, OpenAI, OpenRouter, and Gemini, giving users the flexibility to choose models that align with their specific needs. Furthermore, the assistant incorporates web research functionalities, allowing it to gather current information from the internet to improve its task performance and understanding. Users can engage with the agent through an interactive chat mode, which makes it easy to pose questions or redirect tasks as desired. In addition, RA.Aid can work in conjunction with 'aider' by using the '--use-aider' command, which enhances its code editing capabilities. It is also equipped with a human-in-the-loop feature, allowing the agent to request user input during task execution to achieve greater precision. By combining automation with human oversight, RA.Aid aims to create a more effective development experience for users.

Activepieces is an intuitive, open-source automation platform that enables teams to build powerful AI-driven workflows without any coding. With 280+ pre-built automation pieces (MCPs), users can easily integrate various applications, streamline repetitive tasks, and automate business processes. The platform offers no-code tools for creating chat interfaces, automating approvals, and generating AI-powered agents. Whether for small businesses or large corporations, Activepieces supports decentralized innovation and seamless collaboration, empowering teams to automate daily operations, improve productivity, and unlock the full potential of AI in their workflows.

Llama 4 Behemoth, with 288 billion active parameters, is Meta's flagship AI model, setting new standards for multimodal performance. Outpacing its predecessors like GPT-4.5 and Claude Sonnet 3.7, it leads the field in STEM benchmarks, offering cutting-edge results in tasks such as problem-solving and reasoning. Designed as the teacher model for the Llama 4 series, Behemoth drives significant improvements in model quality and efficiency through distillation. Although still in development, Llama 4 Behemoth is shaping the future of AI with its unparalleled intelligence, particularly in math, image, and multilingual tasks.

Llama 4 Maverick is a cutting-edge multimodal AI model with 17 billion active parameters and 128 experts, setting a new standard for efficiency and performance. It excels in diverse domains, outperforming other models such as GPT-4o and Gemini 2.0 Flash in coding, reasoning, and image-related tasks. Llama 4 Maverick integrates both text and image processing seamlessly, offering enhanced capabilities for complex tasks such as visual question answering, content generation, and problem-solving. The model’s performance-to-cost ratio makes it an ideal choice for businesses looking to integrate powerful AI into their operations without the hefty resource demands.

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.

Nelly serves as an all-in-one AI agent platform that enables users to create, test, share, and deploy AI agents effortlessly, without any coding skills necessary. By utilizing Nelly Studio, individuals can design personalized AI agents by simply providing natural language instructions and formatting them with headings, lists, and various content types. These agents can be enhanced with multiple tools, including a web browser and a database, to effectively perform their designated tasks. Users can tackle complex challenges by breaking them into smaller components, assigning them to specialized sub-agents, which allows for the development of a collaborative team of agents to manage sophisticated workflows. With Nelly, users can engage in natural, fluid conversations with their AI agents, who grasp context and maintain a coherent dialogue, thus removing the necessity for specific commands or syntax. Conversations are systematically organized into threads to improve efficiency and clarity. Furthermore, users have the capability to create departments and arrange their agents through a simple drag-and-drop interface, facilitating the construction of their ideal AI team while enhancing overall productivity. This platform not only streamlines the process of AI interaction but also empowers users to customize their experience to meet their unique needs.

OpenTools serves as an API platform that empowers developers to enhance large language models (LLMs) with dynamic features like web searches, location information, and web scraping, all through a single, cohesive interface. By connecting to a registry of Model-Context Protocol (MCP) servers, OpenTools enables LLMs to utilize various tools without the necessity of separate API keys for each. The platform is designed to be compatible with numerous LLMs, including those facilitated by OpenRouter, and offers robustness against service interruptions, allowing for effortless transitions between different models. Developers can easily invoke tools by making straightforward API calls, where they indicate their preferred model and the tools they wish to use, while OpenTools manages both authentication and execution on their behalf. Remarkably, the service only incurs charges for successful tool executions, featuring a transparent, cost-effective token pricing system that is overseen through a streamlined billing portal. This strategy significantly eases the incorporation of external tools into LLM applications and minimizes the intricacies associated with managing multiple APIs, making it an attractive option for developers seeking efficiency in their projects. Overall, OpenTools represents a pivotal innovation in enhancing the functionality of language models by simplifying access to vital external resources.

GLM-4.6 builds upon the foundations laid by its predecessor, showcasing enhanced reasoning, coding, and agent capabilities, resulting in notable advancements in inferential accuracy, improved tool usage during reasoning tasks, and a more seamless integration within agent frameworks. In comprehensive benchmark evaluations that assess reasoning, coding, and agent performance, GLM-4.6 surpasses GLM-4.5 and competes robustly against other models like DeepSeek-V3.2-Exp and Claude Sonnet 4, although it still lags behind Claude Sonnet 4.5 in terms of coding capabilities. Furthermore, when subjected to practical tests utilizing an extensive “CC-Bench” suite that includes tasks in front-end development, tool creation, data analysis, and algorithmic challenges, GLM-4.6 outperforms GLM-4.5 while nearing parity with Claude Sonnet 4, achieving victory in approximately 48.6% of direct comparisons and demonstrating around 15% improved token efficiency. This latest model is accessible through the Z.ai API, providing developers the flexibility to implement it as either an LLM backend or as the core of an agent within the platform's API ecosystem. In addition, its advancements could significantly enhance productivity in various application domains, making it an attractive option for developers looking to leverage cutting-edge AI technology.

Gemini Enterprise app is a comprehensive agentic AI platform designed to improve productivity and collaboration across organizations. It enables users to connect various workplace tools and data sources, providing a unified environment for searching, analyzing, and generating content. The platform supports multi-step automation through AI agents that can perform tasks across different applications without manual intervention. Users can leverage prebuilt Google agents or create custom agents using a no-code interface, making AI accessible to both technical and non-technical teams. Gemini Enterprise app also offers centralized control over data access, permissions, and workflows, ensuring secure and compliant operations. It is suitable for various departments, including marketing, sales, engineering, HR, and finance. By grounding AI outputs in enterprise data, it delivers more accurate and relevant results. Overall, it helps organizations operate more efficiently and make data-driven decisions.

Novelcrafter is an innovative writing platform powered by AI, designed to assist authors at every stage of their storytelling journey, encompassing everything from brainstorming ideas and developing characters to drafting, reviewing, and finalizing their written works. The platform features a specialized “Codex” wiki that allows writers to organize essential elements such as characters, settings, lore, and world-building details, promoting consistency and ease of access. Additionally, it provides various structured planning modes that include acts, chapters, and scenes, enabling authors to transition effortlessly between the planning phase and the writing interface. Authors have the flexibility to utilize customizable AI tools, allowing them to link their own API keys (such as OpenAI, Claude, or local LLMs) and create specific prompts, or they can choose to write manually without any AI assistance. Furthermore, Novelcrafter boasts a distraction-free writing mode, keeps a revision history, supports the import and export of documents in formats like Word, Markdown, and HTML, and offers mobile compatibility for writers who need to jot down ideas while on the move. This platform seeks to empower writers by providing a comprehensive suite of tools tailored to enhance creativity and streamline the writing process.

Scraib.app is a macOS writing assistant powered by AI that resides in the menu bar, allowing users to select text from any application and improve it by pressing Control + R, which enhances grammar, clarity, and style. Users have the flexibility to set custom rules to align with their preferred tone, and unlike other writing software that requires switching between applications, Scraib seamlessly integrates with various platforms, including Slack, Outlook, Pages, Word, Chrome, and Figma. It prioritizes user privacy by offering options to work with different AI providers like ChatGPT, Claude, and others, while also allowing for local operation with supported models, ensuring that sensitive data remains secure. Designed for efficiency, it minimizes workflow interruptions, enabling users to refine their text without leaving their current application, making it an ideal tool for enhancing written communication on the fly. Additionally, Scraib's intuitive shortcut-based system enhances productivity, allowing for quick adjustments and refinements directly where the text exists.

Liquid Apollo is a streamlined mobile application that facilitates completely on-device, cloud-independent AI interactions, allowing users to interact with sophisticated language and vision models in a secure, private manner with minimal delays. It features a collection of compact foundation models sourced from the company's LEAP platform, enabling users to compose messages, send emails, converse with a personal AI assistant, create digital characters, or utilize image-to-text functions, all while maintaining offline capabilities and ensuring no data is transmitted beyond the device. Optimized for immediate responsiveness and offline functionality, Apollo guarantees that all inference occurs locally, eliminating the need for API calls, external servers, or logging of user data. This application acts as both a personal AI exploration tool and a development environment for those utilizing LEAP models, allowing users to effectively assess a model's performance on their specific mobile devices prior to more widespread implementation. Additionally, Apollo's design emphasizes user autonomy, ensuring a seamless experience free from external interruptions or privacy concerns.

The GLM-4.6V is an advanced, open-source multimodal vision-language model that belongs to the Z.ai (GLM-V) family, specifically engineered for tasks involving reasoning, perception, and action. It is available in two configurations: a comprehensive version with 106 billion parameters suitable for cloud environments or high-performance computing clusters, and a streamlined “Flash” variant featuring 9 billion parameters, which is tailored for local implementation or scenarios requiring low latency. With a remarkable native context window that accommodates up to 128,000 tokens during its training phase, GLM-4.6V can effectively manage extensive documents or multimodal data inputs. One of its standout features is the built-in Function Calling capability, allowing the model to accept various forms of visual media — such as images, screenshots, and documents — as inputs directly, eliminating the need for manual text conversion. This functionality not only facilitates reasoning about the visual content but also enables the model to initiate tool calls, effectively merging visual perception with actionable results. The versatility of GLM-4.6V opens the door to a wide array of applications, including the generation of interleaved image-and-text content, which can seamlessly integrate document comprehension with text summarization or the creation of responses that include image annotations, thereby greatly enhancing user interaction and output quality.

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.

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.

GLM-4.5V is an evolution of the GLM-4.5-Air model, incorporating a Mixture-of-Experts (MoE) framework that boasts a remarkable total of 106 billion parameters, with 12 billion specifically dedicated to activation. This model stands out by delivering top-tier performance among open-source vision-language models (VLMs) of comparable scale, demonstrating exceptional capabilities across 42 public benchmarks in diverse contexts such as images, videos, documents, and GUI interactions. It offers an extensive array of multimodal functionalities, encompassing image reasoning tasks like scene understanding, spatial recognition, and multi-image analysis, alongside video comprehension tasks that include segmentation and event recognition. Furthermore, it excels in parsing complex charts and lengthy documents, facilitating GUI-agent workflows through tasks like screen reading and desktop automation, while also providing accurate visual grounding by locating objects and generating bounding boxes. Additionally, the introduction of a "Thinking Mode" switch enhances user experience by allowing the selection of either rapid responses or more thoughtful reasoning based on the situation at hand. This innovative feature makes GLM-4.5V not only versatile but also adaptable to various user needs.

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.

Repo Prompt is an AI coding assistant designed specifically for macOS, which serves as a context engineering tool that empowers developers to interact with and refine codebases through the use of large language models. By enabling users to select particular files or directories, it allows for the creation of structured prompts that contain only the most relevant context, thereby facilitating the review and application of AI-generated code alterations as diffs instead of requiring rewrites of entire files, which ensures meticulous and traceable modifications. Additionally, it features a visual file explorer for efficient project navigation, an intelligent context builder, and CodeMaps that minimize token usage while enhancing the models' comprehension of project structures. Users benefit from multi-model support, enabling them to utilize their own API keys from various providers such as OpenAI, Anthropic, Gemini, and Azure, ensuring that all processing remains local and private unless the user chooses to send code to a language model. Repo Prompt is versatile, functioning as both an independent chat/workflow interface and as an MCP (Model Context Protocol) server, allowing for seamless integration with AI editors, making it an essential tool in modern software development. Overall, its robust features significantly streamline the coding process while maintaining a strong emphasis on user control and privacy.

Knolli serves as an AI copilot platform that allows users to create, deploy, and expand tailored AI copilots and agents without the necessity of coding by converting knowledge, documents, datasets, and proprietary materials into engaging, conversational assistants. This platform features a no-code workspace where individuals, teams, and businesses can articulate their concepts in simple terms, enabling Knolli to automatically organize uploaded materials into a functional AI copilot. Additionally, it ensures data is organized and safeguarded through encrypted private knowledge bases while seamlessly integrating with tools like CRMs, file storage systems, and databases to provide real-time data for contextually relevant interactions. Knolli accommodates a multi-agent framework that allows various specialized agents to operate within a single copilot, offers pre-designed templates for frequent scenarios, and supports custom branding and white-label solutions. Users can also benefit from comprehensive analytics to track performance, usage metrics, and return on investment. Moreover, Knolli enhances productivity by providing workflow automation, which empowers copilots to carry out complex tasks and synchronize with current systems effortlessly. This robust set of features makes Knolli a versatile solution for organizations looking to leverage AI effectively.

TexTab is a productivity application designed for macOS that empowers users to convert AI-driven tasks into instant keyboard shortcuts, facilitating efficient text processing and automation without the need to switch between different applications. It functions at the system level, allowing users to highlight text in any macOS program, such as web browsers, email clients, coding environments, and documents, and execute AI actions with just one keystroke, streamlining tasks like translation, summarization, rewriting, or formalization into easily accessible commands. Users have the flexibility to create an unlimited number of customized AI actions, each with its distinct shortcut, and can connect to various AI providers—including OpenAI, Anthropic, Groq, Perplexity, or OpenRouter—using their personal API keys, ensuring that their data remains confidential and expenses are managed effectively; the API requests are sent directly to the provider without routing through TexTab’s servers. Additionally, the application boasts features such as a one-click AI prompt enhancer, built-in plugins like a pop-up AI chat, a QR code generator, an image converter, and a color picker, all designed to enhance user experience and productivity. This comprehensive suite of tools makes TexTab an invaluable asset for anyone looking to leverage AI capabilities seamlessly within their workflow.

Agent Zero is an innovative open source framework for AI agents that enables the development of autonomous assistants capable of executing intricate tasks through direct interaction with computer systems. This platform offers a unique setting where AI agents can access real system functions, empowering them to run commands, write and execute code, navigate the internet, analyze data, and oversee workflows as part of comprehensive automation solutions. Unlike a standard chat interface, Agent Zero operates within its isolated virtual environment, enabling it to engage with the operating system, install necessary tools, run scripts, and manage tasks across various components seamlessly. The framework prioritizes transparency and developer control, allowing users to monitor, adjust, and personalize agent behavior, tool accessibility, and information processing methods. With a modular architecture, Agent Zero facilitates the dynamic creation and utilization of tools, all while maintaining a consistent memory for enhanced performance. This makes it an ideal choice for developers aiming to build highly customizable and efficient AI-driven workflows.

Octrafic is a command-line tool that leverages AI and is available as open source, aimed at simplifying the process of automated API testing and exploration by allowing users to communicate with APIs in natural language rather than having to write complex scripts or set up intricate testing frameworks. By simply directing the tool to any HTTP API or OpenAPI specification, users can articulate their testing requirements in straightforward English, prompting the integrated AI agent to create test scenarios, perform actual HTTP requests, verify responses, and generate organized results. This tool streamlines the entire testing process, encompassing endpoint discovery, request formulation, schema checks, and error identification, which enables developers to prioritize testing logic without getting bogged down by the underlying implementation specifics. Additionally, it accommodates real-time execution against live APIs, ensuring the accuracy of status codes and behaviors without the need for mock setups, and it can also produce aesthetically formatted PDF reports for effective communication with teams or stakeholders. With its user-friendly approach, Octrafic represents a significant advancement in making API testing more accessible and efficient.

Trinity Large Thinking is an innovative open-source reasoning model crafted by Arcee AI, tailored for intricate, multi-step problem solving and workflows involving autonomous agents that necessitate extended planning and the use of various tools. This model features a sparse Mixture-of-Experts architecture, boasting a remarkable total of around 400 billion parameters, with approximately 13 billion being active for each token, which enhances its efficiency while ensuring robust reasoning capabilities across a range of tasks, including mathematical calculations, code generation, and comprehensive analysis. A notable advancement in this model is its ability to perform extended chain-of-thought reasoning, which allows it to produce intermediate "thinking traces" prior to delivering final solutions, thereby boosting accuracy and reliability in complex situations. Furthermore, Trinity Large Thinking accommodates a substantial context window of up to 262K tokens, allowing it to effectively process lengthy documents, retain context during prolonged interactions, and function seamlessly in continuous agent loops. This model's design reflects a commitment to pushing the boundaries of what automated reasoning systems can achieve.

GPT-5.5 Pro is a next-generation AI model built for execution-heavy tasks across coding, research, business analysis, and scientific workflows. It can interpret complex instructions, break them into steps, and carry work through to completion using tools and automation. The model supports tasks such as generating documents, building applications, analyzing datasets, and navigating software environments. It is designed to operate across tools, enabling seamless workflows from idea to output. In addition, GPT-5.5 Pro integrates with workspace agents—customizable AI agents that automate recurring and multi-step processes across teams. These agents can handle tasks like lead research, reporting, and workflow automation, running independently or on schedules. Built with enterprise-grade safeguards, the model ensures secure and controlled automation. It helps organizations improve productivity by reducing manual effort and accelerating decision-making. GPT-5.5 Pro is ideal for teams looking to scale operations and handle complex workloads efficiently.

Sourcebot is a self-hosted platform for code comprehension that assists developers and AI agents in searching, navigating, and reasoning through extensive codebases, regardless of their size. It allows teams to index repositories from various sources, including GitHub, GitLab, and Bitbucket, enabling exploration through a cohesive interface that offers quick, multi-repository searches with sophisticated filtering options, regex capabilities, and queries tailored to specific programming languages. The platform features an "ask mode" that lets users ask questions in everyday language, while an integrated language model scours the indexed code, tracks references, and provides structured responses with inline citations linked to the corresponding code snippets. Beyond search functionalities, Sourcebot enhances the development experience by incorporating IDE-level navigation tools such as go-to-definition and find-references across all repositories, as well as a built-in file explorer that showcases syntax highlighting and complete visibility of code. This comprehensive set of features empowers developers to work more efficiently and gain deeper insights into their projects.

OllaCoder serves as a private AI coding assistant tailored for VS Code, catering specifically to developers who prefer not to upload their source code to external servers. Operating locally, it utilizes your personal Ollama models and integrates features such as agent mode, inline edits, codebase chat, intelligent autocomplete, MCP servers, and a local-first runtime all within a single editor interface. The core philosophy behind OllaCoder emphasizes the notion that software development is a personal endeavor, asserting that your code should remain under your control while providing an AI assistant that is robust, transparent, and unobtrusive. It primarily communicates with your local Ollama instance, ensuring that prompts, completions, and modifications remain on your device; cloud services are optional, with API keys securely stored in the OS keychain. OllaCoder's agent mode is capable of planning tasks, modifying files, executing terminal commands, and confirming the accuracy of its work, allowing users to approve, reject, or revert any action taken. Additionally, the inline edits feature enables users to select a function, specify the desired change, and examine a real diff change by change, enhancing the coding experience. Overall, OllaCoder represents a significant step forward in maintaining code privacy while providing powerful AI-assisted development tools.

Whim is a cloud-based development workspace designed for deploying AI coding agents with remarkable speed and efficiency. It provides developers the ability to operate AI coding agents such as Claude Code and Codex within isolated cloud containers, rather than relying on local machines. Each assignment is allocated a dedicated sandboxed Ubuntu environment, offering complete shell access, isolated git branches, and real-time terminal streaming. This setup facilitates seamless integration of AI coding agents into the daily operations of developers and teams, promoting parallelism, collaboration, and eliminating the need for local configuration. Users can easily link a repository, compose a prompt, and the AI agent begins its tasks within a protected cloud container that is accessible from any device. Additionally, multiple tasks can be executed at once, enabling experimentation with various strategies, focusing on different features, or allowing an orchestrator to manage a team of agents without interference. Whim also supports native CLI runtimes for Claude and GPT models, with plans to incorporate more models via OpenRouter in the future. This versatility positions Whim as a powerful tool for enhancing productivity in software development environments.

Laguna XS.2 represents Poolside’s innovative open-weight coding model, distinguished as the lightest and quickest member of the Laguna series. This model features a total of 33 billion parameters in a Mixture of Experts setup, with 3 billion parameters activated, and has been meticulously trained in-house using 30 trillion tokens. As the latest generation model accessible to the public, it embodies a second-generation architecture and marks Poolside’s inaugural open-weight offering, drawing from insights gained during the training of Laguna M.1 with synthetic data and reinforcement learning techniques. Specifically designed to enhance agentic coding workflows, Laguna XS.2 excels in coding, acting, and rapidly iterating, particularly within Poolside’s coding agent environment. This model is particularly advantageous for developers and teams seeking a lightweight, efficient coding solution rather than a more cumbersome frontier system. Released under the permissive Apache 2.0 license, it empowers the community to assess, fine-tune, quantize, and build upon its weights, fostering a collaborative development atmosphere. In essence, Laguna XS.2 not only provides a robust platform for agentic coding but also encourages innovation and experimentation among its users.

Laguna M.1 stands out as Poolside's most proficient model for agentic coding, meticulously developed in-house specifically for enhancing software development workflows. This model features a total of 225 billion parameters, utilizing a Mixture of Experts architecture with 23 billion activated parameters, and has been trained entirely within the organization on a dataset consisting of 30 trillion tokens, leveraging the power of 6,144 interconnected NVIDIA H200 GPUs. Poolside undertook the task of training Laguna M.1 from the ground up, employing its proprietary data, dedicated training codebase, and an asynchronous on-policy reinforcement learning approach within its agent framework, all tailored for agentic coding applications. The design of the model ensures optimal performance within Poolside's coding agent, enabling it to effectively reason through software tasks, interact with various tools, edit code, execute tests, and facilitate extended autonomous development sessions. Specifically crafted for developers and teams tackling intricate coding challenges, Laguna M.1 offers enhanced capabilities in reasoning, architectural comprehension, terminal operations, and multi-step execution, surpassing what lighter models can achieve. Ultimately, its robust feature set positions it as an essential asset for those engaged in demanding software projects.

Sakana Fugu is a multi-agent AI platform and AI model that gives users access to coordinated model intelligence through one API. Instead of relying on one frontier model, Fugu dynamically selects, routes, and coordinates multiple expert models to complete complex tasks more effectively. The system is based on research into learned model orchestration, including the TRINITY and Conductor approaches for assembling agents and guiding collaboration patterns. Fugu is designed for coding, code review, reasoning, research, paper reproduction, cybersecurity analysis, patent investigation, and other work that benefits from multiple specialized agents. Users can access Fugu and Fugu Ultra through an OpenAI-compatible API, making integration easier for existing workflows and developer tools. Fugu is positioned as the default option for everyday use because it balances performance and latency. Fugu Ultra is built for difficult, high-value tasks where maximum quality matters more than speed. The platform also gives organizations the ability to opt out of specific models or providers for data, privacy, compliance, or internal policy reasons. Sakana Fugu helps users reduce dependence on a single AI vendor while gaining a flexible orchestration layer for advanced multi-step AI work.

YeeroAI serves as a comprehensive AI knowledge platform that transforms conversations into enduring knowledge and cultivates a network of ideas. Each dialogue contributes to a personal repository of wisdom, empowering users to delve into various concepts, juxtapose models like GPT, Claude, and Gemini, and develop an AI memory that enhances its intelligence over time. The platform treats every message as a foundational element of a user's knowledge base, with each conceptual branch serving to broaden their thought processes. By automatically identifying key insights from discussions, YeeroAI constructs vector indexes and integrates pertinent context into future dialogues, ensuring that the knowledge base becomes increasingly beneficial with each user interaction. Its innovative Git-style branch management system enables individuals to fork, merge, and revisit their thought lines seamlessly, preventing any loss of direction. Additionally, the ability to engage multiple leading AI models in parallel chats allows for simultaneous inquiries and side-by-side answer comparisons. Furthermore, YeeroAI offers comprehensive management of the entire AI application lifecycle, enabling users to articulate ideas in straightforward language, create HTML applications, and enhance them through AI-driven refinement, thus fostering a creative and iterative development environment. The platform truly transforms the way users engage with knowledge and AI technology.

OpenRouter Fusion transforms a prompt into a compact deliberation process involving multiple models, allowing users to access combined results as effortlessly as they would from a single model. A consortium of specialized models examines the prompt simultaneously while utilizing web search and web fetch capabilities, after which a judge model evaluates their outputs and presents a structured analysis featuring consensus, contradictions, partial coverage, unique insights, and blind spots. This comprehensive analysis culminates in the final answer, enabling users to gain insights from various viewpoints instead of depending solely on one model. Fusion is particularly advantageous in scenarios where a single model falls short, such as in research, expert evaluations, comparative prompts, multi-domain inquiries, or any situation where inaccuracies could be costly. Users have the flexibility to access Fusion directly via the openrouter/fusion model alias, activate it as a fusion server tool, or set it up through the Fusion plugin; all these methods utilize the same underlying framework. By providing these versatile entry points, Fusion caters to a wide range of user needs and preferences.

Wafer is revolutionizing enterprise AI by offering the quickest open-source LLMs, enabling serverless and dedicated inference designed specifically for production workloads. With its serverless inference, teams can utilize top-tier open models without the burden of infrastructure and deployment challenges, providing rapid APIs that include GLM-5.2-Fast for reduced latency through EAGLE speculative decoding and a guaranteed throughput SLA, alongside GLM-5.2, which serves as a flagship model boasting enhanced coding and reasoning abilities. Wafer's innovative technology employs agents to optimize inference throughout the stack, pinpointing and addressing bottlenecks in orchestration, algorithms, serving engines, GPU kernels, and various hardware setups. This system meticulously profiles the stack to determine whether latency or throughput issues arise from factors such as scheduling, decoding, kernels, memory pressure, or hardware compatibility, and then it explores numerous paths to deliver the most effective solution. Rather than depending on a singular switch or heuristic, Wafer undertakes a comprehensive search of combinations involving models, engines, kernels, and hardware to maximize performance. By continually refining these combinations, Wafer ensures that enterprises can operate at peak efficiency while leveraging the best of open-source technologies.

Minds allows you to establish your own AI agent in less than a minute. This always-available assistant can be accessed through Telegram or email without needing any applications, coding, or installations. Minds efficiently manages tasks such as scheduling, email handling, research, reminders, daily chores, appointments, travel plans, side ventures, sales, wellness activities, family scheduling, and much more, enabling you to simply state your requirements and let your Mind take care of the rest. Yet, one Mind is merely the starting point: it can be utilized by multiple users, connect with other Minds for collaboration, and expand into a specialized team of Minds as tasks become more complex. The Bazaar offers one-click Minds like a General Assistant, designed to handle bookings, reminders, and everyday tasks, while integrating with tools such as Google Calendar, Gmail, Google Tasks, Google Sheets, and Google Docs. Additionally, creators have the ability to develop Skills by articulating desired outcomes in straightforward language, further enhancing the utility of the Minds platform. With these innovative features, Minds is revolutionizing how we interact with technology to streamline our daily lives.

AI Workers are independent agents crafted specifically for your organization; these advanced AI entities function like genuine colleagues rather than mere chatbots awaiting instructions. Each comes equipped with its own professional identity, including an email, phone number, and presence on platforms like Slack and Teams, demonstrating initiative and the capacity to operate around the clock without needing prompts. Rather than dictating every detail of their tasks, you simply define the objectives, and they autonomously develop the necessary workflows, accommodating a variety of tasks such as generating daily reports, conducting weekly follow-ups, managing CRM updates, handling client interactions, performing research, overseeing content creation, executing finance responsibilities, coordinating HR activities, supporting design projects, and assisting with development operations. This system includes a diverse range of specialized AI Workers tailored for functions in marketing, development, design, HR, finance, and more, with each worker crafted for a specific role and practical applications. Furthermore, AI Workers can integrate seamlessly with over 3,000 tools, such as Slack, Microsoft Teams, Gmail, Notion, HubSpot, Salesforce, and various other business applications, ensuring they fit effortlessly into your existing workflows. Their versatility and adaptability make them an invaluable asset for enhancing productivity and streamlining processes in any business environment.

Siket is an innovative new tab workspace that transforms your browser's new tab into a centralized command hub for managing todos, focus, emails, calendar events, tab sessions, habits, and AI features. Designed to prioritize keyboard navigation and optimized for Chrome, Edge, and Brave browsers, it consolidates the various aspects of your daily tasks into a single, cohesive starting point. Rather than opening a blank tab and getting sidetracked by other applications, users can initiate their workflow through Siket, allowing them to quickly identify urgent tasks, organize their day, manage emails, monitor habits, restore previous work sessions, and dive into focused work with ease. The platform offers practical guides tailored to real-world work scenarios, such as morning routines, email management, deep focus periods, Slack communications, research activities, and accessing cloud files. By seamlessly integrating todos, calendar events, habits, and emails into a unified schedule, Siket aids users in transforming their fragmented commitments into a more streamlined daily agenda. For tasks involving research and concentrated effort, the Focus feature can be effectively combined with Tab Sessions, ensuring that research tabs are restored after each session, thereby maintaining context and enhancing productivity. This holistic approach not only simplifies task management but also fosters a more organized and efficient work environment.

LLMeter is a comprehensive open-source platform designed for monitoring AI costs, allowing developers to manage their expenditures across various providers like OpenAI, Anthropic, DeepSeek, OpenRouter, Mistral, and Azure OpenAI from a single dashboard. By simply connecting read-only provider keys, teams can instantly access detailed insights into actual costs, daily usage trends, model-specific analytics, and potential areas for optimization, all within approximately 30 seconds and without any need for SDK installation, endpoint modifications, or rerouting production traffic through a proxy. Since it facilitates direct communication with model providers, LLMeter introduces no additional latency, avoids becoming a single point of failure, and does not access or store any user prompts or completions. Additionally, budget alerts notify teams prior to exceeding their daily or monthly spending thresholds, while anomaly detection features help catch unexpected usage surges before they escalate. The platform's dashboard provides a clear overview of the costs associated with various providers, models, endpoints, customers, and environments, and its integration with OpenRouter enhances transparency by covering over 500 models, ensuring users have a robust tool for managing their AI-related expenditures efficiently. Ultimately, LLmeter empowers teams to make informed financial decisions regarding their AI usage.

Dograh is a self-hostable voice agent platform that is open source and features a no-code workflow builder designed for developing production-ready voice agents. Teams have the flexibility to select their preferred inbound channels, speech-to-text services, language models, text-to-speech options, and telephony providers, or they can opt for innovative speech-to-speech models that facilitate direct audio interactions with seamless turn-taking, interruption management, and minimal latency. The platform caters to both inbound and outbound calling, offering widgets, telephony integrations, observability, tracing capabilities, real-time analytics, and a hybrid approach that combines pre-recorded voice with TTS, all while supporting over 70 languages. Additionally, the MCP server enables various agent runtimes, including Claude Code, Cursor, OpenClaw, and Codex, to create, modify, and deploy voice agents directly from development environments. Dograh can be operated on personal servers, within a private cloud or virtual private cloud, or in a managed setting, ensuring that models can be hosted entirely within the user's infrastructure. With its extensive features and adaptability, Dograh stands out as a versatile solution for teams looking to innovate in voice technology.

ZGI is an open-source AI platform designed for enterprises to create business-oriented agents that leverage their specific data, tools, workflows, models, and expertise. With its Agent Runtime feature, agents can efficiently load Skills, access real-time company insights, utilize various tools, and quickly deliver valuable results. The Model Gateway integrates multiple global and domestic model providers, such as OpenAI, Anthropic, Google, DeepSeek, and Qwen, enabling teams to select the most suitable model for each agent based on criteria like quality, availability, cost, or geographic location while maintaining centralized control over access, quotas, routing policies, and fallback options. The comprehensive product workspace encompasses the entire agent development lifecycle: Agent Studio merges Skills, knowledge, tools, and models; Workflows facilitate complex, multi-step processes and allow users to monitor each execution closely; the Database enables natural-language queries over real-time business information while adhering to access governance; Model Management oversees provider connections and enterprise-level routing; and Knowledge Assets transform company documents into searchable, context-aware knowledge repositories. By integrating these features, ZGI aims to streamline the deployment and management of AI agents within enterprises.

Assistable is a comprehensive AI conversation platform that facilitates the creation, testing, deployment, and monitoring of hybrid AI agents capable of responding to calls, texts, WhatsApp messages, and web chats while maintaining a unified memory across all channels. These agents effectively manage entire customer dialogues, interpret intent, retrieve information, qualify leads, and perform tasks such as booking, rescheduling, or canceling appointments, as well as creating support tickets, updating contact details, adding notes and tasks, and escalating issues to human agents when necessary. Thanks to the continuity of memory across channels, a customer can initiate a conversation via phone and seamlessly transition to text or chat later without needing to repeat any previous exchanges. Users can easily outline the desired functionalities of an agent using simple English, connect Assistable through MCP, and have the platform automatically generate the agent, attach relevant tools, connect various communication channels, and designate a phone number. Furthermore, these agents can access real-time calendar availability, follow up with potential leads, and facilitate both incoming and outgoing conversations, ensuring a seamless customer experience throughout the interaction process. This versatility ensures that businesses can efficiently manage customer inquiries while maximizing engagement across different communication platforms.

VibeView is a cloud-based simulator accessible through browsers on iOS, Android, Apple TV, and Android TV, featuring real-time connection capabilities, AI-driven mobile app testing, an SDK that can be embedded into devices, and the ability to generate instantly shareable app previews. This innovative tool empowers developers, QA teams, and product managers to execute mobile applications directly from their browsers, eliminating the need for a Mac, Xcode, Android Studio, or physical devices. Users can simply upload their iOS or Android builds and within moments, stream a live simulator, sharing links that allow teammates, clients, beta testers, or sales prospects to engage with the app immediately—without requiring any installations, TestFlight, or provisioning profiles. Additionally, VibeView serves as a versatile development tool as well, where executing the command vibeview dev allows your React Native application to stream to a browser simulator with complete hot reload capabilities—enabling immediate feedback on code edits made locally, all without the necessity of a Mac, Xcode, Android Studio, or a physical device. As a result, both testing and development processes become streamlined and more efficient, enhancing collaboration and productivity.

MiMo-V2.6-Pro-UltraSpeed is Xiaomi MiMo’s accelerated serving option for MiMo-V2.6-Pro, built for applications that require very high output speed without changing the underlying model quality. Xiaomi states that UltraSpeed can generate output at up to 20 times the speed of the standard MiMo-V2.6-Pro configuration. The model retains MiMo-V2.6-Pro’s natively omnimodal capabilities across coding, agentic workflows, visual reasoning, computer use, and research. Developers can use it for long-horizon software engineering, automation, debugging, tool-driven tasks, and other workloads that benefit from rapid model responses. Its multimodal abilities also support frontend generation, presentation creation, 3D modeling, interactive environments, and visual feedback loops. The broader MiMo-V2.6 architecture combines coding capabilities with 3D spatial reasoning, multimodal perception, and computer-use agent functionality. Xiaomi positions UltraSpeed for real-time interaction and other workflows where response latency is especially important. The accelerated model is offered through MiMo Desktop and can also be called through the Xiaomi MiMo API Platform. MiMo-V2.6-Pro-UltraSpeed is intended for developers and organizations that prioritize maximum generation speed while retaining the capabilities of Xiaomi’s higher-end MiMo-V2.6-Pro model.

Holo4 is H Company's series of generalist computer-use and agentic AI models built to perform multi-step work across software interfaces. It is available as Holo4 27B, a dense 27-billion-parameter model, and Holo4 35B-A3B, a Mixture-of-Experts model containing 35 billion total parameters with 3 billion active. Holo4 can interact with applications by clicking and typing through graphical interfaces, writing and executing code, or calling MCP and API tools. The same model can operate across desktops, websites, Android devices, code sandboxes, and business APIs without requiring developers to select a separate specialized model for each environment. H Company trained Holo4 using 127 billion supervised fine-tuning tokens, with approximately three-quarters consisting of successful agentic trajectories spanning desktop, web, MCP/API, and mobile tasks. Reinforcement learning then trained separate experts for desktop and web interaction and for terminal, MCP, and API work before merging them into a single model. Holo4 27B scored 85.2% on OSWorld, 61.7% on OSWorld 2.0, 45.4% on AutomationBench, and 85.1% on AndroidWorld in the evaluations reported by H Company. The models support a 256K context window, with the 27B model positioned for greater accuracy on long multi-step tasks and the 35B-A3B version positioned as a faster and less expensive alternative. Holo4 is available through a hosted API and downloadable model weights, enabling developers and enterprises to build agents that perform workflows spanning multiple applications and interaction methods.

d1 represents the inaugural decision model from Liquid AI, designed to provide answers to structured inquiries regarding a scenario through probabilities instead of generating textual responses. This model processes unstructured data, including text and images, alongside one or more questions in a single streamlined forward pass, delivering probabilities without producing output tokens. It accommodates three distinct question formats: noul for binary yes-or-no decisions, choice for selecting from various labels with associated probabilities, and score for assessing an input along an ordered scale using probability-based levels. Furthermore, d1 can evaluate multiple questions simultaneously within a single request, enhancing its utility in real-time decision-making processes. Liquid AI showcases d1 in various applications, including filtering support tickets, navigating code repositories, categorizing documents into intelligent folders, managing websites, condensing agent context, inspecting industrial components, and even engaging in gameplay from screenshots, illustrating its versatile capabilities. This model's ability to handle diverse data types and provide rapid responses indicates a significant advancement in decision-making technology.

StackStats is a powerful desktop analytics tool designed specifically for Substack writers and independent publishers. It efficiently imports or syncs newsletter data, offering a comprehensive overview of subscriber growth, engagement levels, churn-risk indicators, audience demographics, acquisition sources, post performance metrics, and growth projections. This allows writers to assess reader engagement more effectively, pinpoint highly active or potentially at-risk subscribers, and analyze the impact of individual posts on overall growth, as well as investigate the times when readers are most likely to open their emails. The analytics are processed locally on the user's computer, and the application is compatible with both Mac and Windows operating systems. Additionally, optional AI features offer users the ability to access summaries, strategic recommendations, actionable plans, and data-driven conversations via a local Ollama model or supported third-party AI services, utilizing the user's own API key. StackStats is available for purchase with a one-time license, ensuring that the app remains functional even after the included update period has expired, providing ongoing value to its users. This makes it an indispensable tool for anyone looking to enhance their publishing efforts and optimize their newsletter performance.

HyCanvas is an online platform that enables users to create a variety of visual content, such as graphics, presentations, videos, whiteboards, documents, spreadsheets, and print designs. The platform features an editing interface that integrates drawing tools, text options, templates, media libraries, branding kits, and export functionalities all within a collaborative environment. Users have the capability to develop slide presentations complete with presenter controls, edit videos using a multitrack timeline, and utilize whiteboards for creative sessions and workshops. Additionally, it offers optional AI tools that can assist in generating presentations and executing design tasks, provided the workspace is linked to its own set of provider credentials. HyCanvas can either be accessed as a cloud-based service or set up for self-hosting, allowing clients to maintain control over their design assets. Projects created within HyCanvas can be exported in popular file formats or stored in its proprietary design-file format, ensuring flexibility for future edits or transfers. This versatility makes HyCanvas an invaluable tool for individuals and teams looking to streamline their creative processes.

RichAutomate is an online platform designed specifically for businesses utilizing WhatsApp to enhance their customer engagement, marketing efforts, sales processes, and support services. It features a collaborative inbox for teams, the ability to send campaign messages using pre-approved templates, contact labeling, and detailed reports on message delivery and read statuses. Additionally, it offers automated responses through keyword recognition, a visual flow builder for creating automated dialogues, and in-chat forms to streamline interactions. Optional AI capabilities enable the system to respond based on a database of business knowledge and transfer conversations to human representatives when necessary. RichAutomate integrates seamlessly with the Meta WhatsApp Cloud API and provides developer API endpoints, webhooks, and documentation for connections with platforms like Zoho CRM, Razorpay, Google Sheets, Shopify, and WooCommerce. Teams can select between monthly subscription plans or billing based on their usage, and a 14-day trial period is offered for new users. The platform is particularly beneficial for direct-to-consumer (D2C) brands, e-commerce businesses, agencies, and organizations in sectors such as education, healthcare, and finance, all of which rely on effective management of customer interactions via WhatsApp. Ultimately, RichAutomate aims to simplify and enhance the way businesses communicate with their customers through this popular messaging service.

Mercury 3 is an innovative generative AI model offered by Inception, designed specifically for developers creating interactive applications. It employs a diffusion-based methodology to generate text concurrently, while also enabling streaming responses, tool utilization, structured outputs, and multimodal inputs that encompass text, images, and audio. By leveraging the Inception API, teams can implement it in various applications, such as conversational assistants, voice-driven platforms, search and browsing functionalities, tutoring solutions, and other responsive tasks that require dynamic generation. This powerful model is accessible through Inception's API and partnered inference providers, featuring compatibility with OpenAI's API for seamless integration into existing applications. Developers can assess the quality of the output and its latency before incorporating the model into their software, utilizing the comprehensive documentation provided with the API and SDK ecosystem. The service operates on a token-based usage system, offering a complimentary allocation along with paid rates for input and output tokens; enterprise-level options provide dedicated capacity and specialized support, ensuring flexibility and scalability for diverse projects. Ultimately, Mercury 3 empowers developers to harness advanced AI capabilities in their applications, streamlining the development process and enhancing user experiences.

xliff.ai serves as a web-based computer-assisted translation tool designed for freelance translators, reviewers, localization teams, and agencies engaged with various translation file formats. It allows users to open and modify XLIFF 1.2/2.0, SDLXLIFF, MQXLIFF, PO, XLSX, and CSV files directly on their local machines without the need for uploading, and it can export the modified content back into its original format. Translators can efficiently translate and assess segments, conduct 15 types of quality checks, and integrate translation memories and glossaries. The platform also offers features like concordance and find-and-replace options, along with the ability to export TMX files, bilingual or review workbooks, QA reports, and pseudo-translated documents. Additionally, the editor comes equipped with keyboard layouts tailored for Trados Studio and memoQ, along with helpful tools such as autosave and undo functionalities. For those needing AI assistance, there's an option for AI review and translation, requiring a provider API key, which ensures that text is sent to the chosen provider only during the active AI process. After installation in a compatible desktop browser, the editor can function offline, providing flexibility for users who may not always have internet access. This makes xliff.ai a versatile choice for professionals in the translation industry looking to enhance their workflow.