Open Source ChromeOS Artificial Intelligence Software - Page 4
Sort By:
Artificial Intelligence Software for ChromeOS
-
Gemini 3 and 200+ AI Models on One PlatformBuild generative AI apps with Vertex AI. Switch between models without switching platforms.
-
Try Google Cloud Risk-Free With $300 in CreditUse your credit across every product. Compute, storage, AI, analytics. When it runs out, 20+ products stay free. You only pay when you choose to.
-
1
Agent Starter Pack
Ship AI Agents to Google Cloud in minutes, not months
Agent Starter Pack is a production-focused framework that provides pre-built templates and infrastructure for rapidly developing and deploying generative AI agents on Google Cloud. It is designed to eliminate the complexity of moving from prototype to production by bundling essential components such as deployment pipelines, monitoring, security, and evaluation tools into a single package. Developers can create fully functional agent projects with a single command, generating both backend and frontend structures along with deployment-ready configurations. The framework supports multiple agent architectures, including ReAct, retrieval-augmented generation, and multi-agent systems, allowing flexibility across use cases. It integrates tightly with Google Cloud services like Vertex AI, Cloud Run, and Terraform-based infrastructure provisioning, enabling scalable and reliable deployments. -
2
Browserbase MCP Server
Allow LLMs to control a browser with Browserbase and Stagehand
Browserbase MCP Server is a server implementation of the Model Context Protocol (MCP) that enables large language models to interact with web browsers programmatically through cloud-based automation. The project provides a standardized interface for connecting AI systems to real-world web environments, allowing them to navigate pages, extract structured data, and perform user-like actions such as clicking, typing, and form submission. It leverages Browserbase infrastructure along with Stagehand to deliver high-performance browser automation with improved speed and efficiency through caching and optimized execution pipelines. The system supports multiple AI models and integrates seamlessly into agent workflows, making it suitable for applications such as web scraping, testing, and intelligent automation. It also includes advanced capabilities such as screenshot capture, DOM analysis, and session persistence, enabling complex interactions across multiple browsing sessions. -
3
CodeMachine
CLI tool for multi-agent workflows and automated code generation
CodeMachine CLI is a command-line orchestration engine designed to run coordinated multi-agent workflows locally. It enables developers to transform high-level specifications into production-ready code by managing planning, architecture, implementation, testing, and validation within a unified environment. CodeMachine CLI supports parallel execution through multiple specialized agents, allowing faster development cycles and scalable automation. Built for flexibility, it can handle anything from simple scripts to complex, long-running workflows that span hours or days. CodeMachine also integrates with various AI engines, assigning roles such as planning, coding, and review to different models for efficient collaboration. -
4
Colab-MCP
An MCP server for interacting with Google Colab
Colab-MCP is an open-source Model Context Protocol server developed by Google that enables AI agents to directly interact with and control Google Colab environments programmatically, transforming Colab into a fully automated, agent-accessible workspace. Instead of relying on manual notebook usage, the system allows MCP-compatible agents to execute code, manage files, install dependencies, and orchestrate entire development workflows within Colab’s cloud infrastructure. This approach bridges the gap between local AI agents and remote high-performance compute environments, allowing users to offload heavy workloads such as machine learning training, data analysis, and dependency-heavy tasks to Colab’s GPU and TPU resources. By exposing Colab as an MCP server, the tool enables seamless integration with a wide range of AI assistants and agent frameworks, creating a standardized interface for tool use and execution. -
MongoDB Atlas runs apps anywhereMongoDB Atlas gives you the freedom to build and run modern applications anywhere—across AWS, Azure, and Google Cloud. With global availability in over 115 regions, Atlas lets you deploy close to your users, meet compliance needs, and scale with confidence across any geography.
-
5
DeepProve
Framework to prove inference of ML models blazingly fast
DeepProve is an advanced cryptographic framework designed to verify machine learning model inference using zero-knowledge proofs, enabling trustless validation of AI computations without exposing underlying data. The project focuses on zkML, a rapidly emerging field that combines machine learning with zero-knowledge cryptography to ensure both privacy and correctness. It supports neural network architectures such as multilayer perceptrons and convolutional neural networks, allowing developers to prove that a model’s output is correct without revealing inputs or model details. deep-prove leverages advanced proof systems such as sumcheck protocols and GKR-based constructions to achieve significantly faster proving times compared to earlier approaches. This makes it viable for real-world applications in industries like healthcare, finance, and blockchain, where sensitive data must remain confidential. -
6
Dia
A TTS model capable of generating ultra-realistic dialogue
Dia is a neural text-to-speech model designed specifically for generating ultra-realistic dialogue in a single pass. Instead of focusing on isolated sentences or flat narration, it is optimized for conversational audio, complete with natural turn-taking, prosody, and pacing. The model can be conditioned on a reference audio sample, allowing you to control emotion, tone, and other stylistic aspects of the speech. It can also produce nonverbal vocalizations like laughter, coughs, clearing the throat, and similar sounds, which are crucial for making synthetic conversations feel human. Dia is released with pretrained checkpoints and inference code, with weights hosted on Hugging Face, so researchers and developers can quickly try it or integrate it into pipelines. The base model currently targets English and has around 1.6 billion parameters, offering a strong balance between realism and computational cost, while the ecosystem also includes Dia2. -
7
Docker Agent
AI Agent Builder and Runtime by Docker Engineering
Docker Agent is an open-source multi-agent runtime developed by Docker that enables developers to define, run, and orchestrate AI agents using simple declarative configuration files instead of traditional code-heavy approaches. It introduces a YAML-based configuration model where users describe agent behavior, tools, models, and interaction logic in a single file, significantly reducing complexity in building AI systems. The runtime supports multi-agent collaboration, allowing specialized agents to delegate tasks to each other and operate as coordinated systems rather than isolated units. It is provider-agnostic, meaning it can integrate with multiple AI model providers such as OpenAI, Anthropic, and local inference engines, helping avoid vendor lock-in. cagent also supports the Model Context Protocol, enabling seamless integration with external tools, APIs, and services. -
8
Dynamiq
An orchestration framework for agentic AI and LLM applications
Dynamiq is an open-source orchestration framework designed to streamline the development of generative AI applications that rely on large language models and autonomous agents. The framework focuses on simplifying the creation of complex AI workflows that involve multiple agents, retrieval systems, and reasoning steps. Instead of building each component manually, developers can use Dynamiq’s structured APIs and modular architecture to connect language models, vector databases, and external tools into cohesive pipelines. The framework supports the creation of multi-agent systems where different AI agents collaborate to solve tasks such as information retrieval, document analysis, or automated decision making. Dynamiq also includes built-in support for retrieval-augmented generation pipelines that allow models to access external documents and knowledge bases during inference. -
9
EnvPool
C++-based high-performance parallel environment execution engine
EnvPool is a fast, asynchronous, and parallel RL environment library designed for scaling reinforcement learning experiments. Developed by SAIL at Singapore, it leverages C++ backend and Python frontend for extremely high-speed environment interaction, supporting thousands of environments running in parallel on a single machine. It's compatible with Gymnasium API and RLlib, making it suitable for scalable training pipelines. -
Auth0 B2B Essentials: SSO, MFA, and RBAC Built InAuth0's B2B Essentials plan gives you everything you need to ship secure multi-tenant apps. Unlimited orgs, enterprise SSO, RBAC, audit log streaming, and higher auth and API limits included. Add on M2M tokens, enterprise MFA, or additional SSO connections as you scale.
-
10
FasterTransformer
Transformer related optimization, including BERT, GPT
FasterTransformer is a high-performance inference library designed to accelerate transformer-based models such as BERT, GPT, and T5 on NVIDIA GPUs. It provides optimized implementations of transformer encoder and decoder layers using CUDA, cuBLAS, and custom kernels to maximize throughput and minimize latency. The library supports multiple deep learning frameworks, including TensorFlow, PyTorch, and Triton, allowing developers to integrate it into existing pipelines without major changes. It includes advanced optimization techniques such as mixed precision, tensor parallelism, and efficient memory management, enabling large models to run across multiple GPUs and nodes. FasterTransformer is particularly focused on inference workloads, where it significantly improves performance compared to standard framework implementations. Although development has transitioned toward TensorRT-LLM, the project remains an important reference for understanding optimized transformer execution. -
11
GLM-4.1V
GLM-4.6V/4.5V/4.1V-Thinking, towards versatile multimodal reasoning
GLM-4.1V — often referred to as a smaller / lighter version of the GLM-V family — offers a more resource-efficient option for users who want multimodal capabilities without requiring large compute resources. Though smaller in scale, GLM-4.1V maintains competitive performance, particularly impressive on many benchmarks for models of its size: in fact, on a number of multimodal reasoning and vision-language tasks it outperforms some much larger models from other families. It represents a trade-off: somewhat reduced capacity compared to 4.5V or 4.6V, but with benefits in terms of speed, deployability, and lower hardware requirements — making it especially useful for developers experimenting locally, building lightweight agents, or deploying on limited infrastructure. Given its open-source availability under the same project repository, it provides an accessible entry point for testing multimodal reasoning and building proof-of-concept applications. -
12
GLM-4.5V
GLM-4.6V/4.5V/4.1V-Thinking, towards versatile multimodal reasoning
GLM-4.5V is the preceding iteration in the GLM-V series that laid much of the groundwork for general multimodal reasoning and vision-language understanding. It embodies the design philosophy of mixing visual and textual modalities into a unified model capable of general-purpose reasoning, content understanding, and generation, while already supporting a wide variety of tasks: from image captioning and visual question answering to content recognition, GUI-based agents, video understanding, and long-document interpretation. GLM-4.5V emerged from a training framework that leverages scalable reinforcement learning (with curriculum sampling) to boost performance across tasks ranging from STEM problem solving to long-context reasoning, giving it broad applicability beyond narrow benchmarks. When it was released, it achieved state-of-the-art results on a large collection of public multimodal benchmarks for open-source models. -
13
Google Research
This repository contains code released by Google Research
Google Research is a massive monorepo that hosts a wide range of research code released by Google Research teams across machine learning, artificial intelligence, robotics, natural language processing, and other advanced domains. Rather than being a single framework, the repository serves as a centralized collection of experimental projects, reference implementations, and reproducible research artifacts. It is intended primarily for researchers and advanced practitioners who want to explore cutting-edge techniques directly from the teams that developed them. The repository includes datasets, training scripts, and model implementations that support academic study and applied experimentation. Because of its breadth, users typically clone only the subdirectories relevant to their specific research interests. Overall, google-research functions as a living archive of state-of-the-art research code supporting both academic and industrial AI innovation. -
14
Hugging Face - Speech To Speech
Open speech-to-speech models and pipelines by Hugging Face toolkit AI
This project from Hugging Face focuses on enabling direct speech-to-speech processing using modern machine learning models. It provides tools and reference implementations that allow audio input to be transformed into audio output without requiring an intermediate text representation. Hugging Face - Speech To Speech builds on recent advances in speech modeling, combining components such as speech recognition, translation, and synthesis into unified pipelines. It is designed to help researchers and developers experiment with multilingual and cross-lingual voice applications. It integrates with the broader Hugging Face ecosystem, making it easier to load pretrained models and run inference. It also serves as a foundation for building real-time or batch audio transformation systems. Overall, it highlights an emerging approach to voice technology that reduces latency and preserves more of the original speech characteristics. -
15
LLM CLI
Access large language models from the command-line
A CLI utility and Python library for interacting with Large Language Models, both via remote APIs and models that can be installed and run on your own machine. -
16
OpenMemory
Local long-term memory engine for AI apps with persistent storage
OpenMemory is a self-hosted memory engine designed to provide long-term, persistent storage for AI and LLM-powered applications. It enables developers to give otherwise stateless models a structured memory layer that can store, retrieve, and manage contextual information over time. OpenMemory is built around a hierarchical memory architecture that organizes data into semantic sectors and connects them through a graph-based structure for efficient retrieval. It supports multiple embedding strategies, including synthetic and semantic embeddings, allowing developers to balance speed and accuracy depending on their use case. OpenMemory integrates with various AI tools and environments, offering SDKs and APIs that simplify adding memory capabilities to applications. OpenMemory also includes features like memory decay, reinforcement, and temporal filtering to ensure relevant information remains prioritized while outdated data gradually loses importance. -
17
Qwen3-ASR
Qwen3-ASR is an open-source series of ASR models
Qwen3-ASR is an automatic speech recognition system in the QwenLM family, developed to convert spoken language into text with strong accuracy and real-time performance. As a specialized ASR variant of the broader Qwen language model ecosystem, it focuses on capturing reliable transcriptions from audio sources such as recordings, live streams, or conversational inputs while supporting low latency use cases. The architecture combines advanced neural acoustic modeling with context-aware language prediction so that outputs maintain both fidelity to the original speech and grammatical coherence. This makes Qwen3-ASR suitable for voice-driven applications like AI assistants, dictation tools, speech analytics pipelines, and accessibility features, where accurate and fluid transcription is critical. -
18
Segment Anything
Provides code for running inference with the SegmentAnything Model
Segment Anything (SAM) is a foundation model for image segmentation that’s designed to work “out of the box” on a wide variety of images without task-specific fine-tuning. It’s a promptable segmenter: you guide it with points, boxes, or rough masks, and it predicts high-quality object masks consistent with the prompt. The architecture separates a powerful image encoder from a lightweight mask decoder, so the heavy vision work can be computed once and the interactive part stays fast. A bundled automatic mask generator can sweep an image and propose many object masks, which is useful for dataset bootstrapping or bulk annotation. The repository includes ready-to-use weights, Python APIs, and example notebooks demonstrating both interactive and automatic modes. Because SAM was trained with an extremely large and diverse mask dataset, it tends to generalize well to new domains, making it a practical starting point for research and production annotation tools. -
19
Super Magic
All-in-one AI productivity platform with agents, workflows, and IM
Magic is an open source all-in-one AI productivity platform designed to help organizations build, deploy, and scale AI-driven applications efficiently. It is not a single tool but a complete product ecosystem composed of multiple integrated systems that work together to enhance productivity across different business scenarios. Magic centers around a general-purpose AI agent system called Super Magic, which can autonomously understand tasks, plan actions, execute workflows, and perform error correction. Alongside this, Magic includes a visual workflow engine that enables users to design complex AI processes using a drag-and-drop interface without requiring extensive coding knowledge. It also provides an enterprise-grade instant messaging system that integrates AI conversations with internal communication, allowing teams to collaborate while leveraging intelligent assistants. Its architecture is built using a microservices approach with containerized services. -
20
Zvec
A lightweight, lightning-fast, in-process vector database
Zvec is an open-source, lightweight, in-process vector database designed to embed directly into applications and serve fast similarity search workloads without the overhead of a separate server process. Developed by Alibaba’s Tongyi Lab, it positions itself as the “SQLite of vector databases” by being easy to integrate, minimal in dependencies, and capable of handling high throughput with low latency on edge devices or small systems. Zvec excels at approximate nearest neighbor search and retrieval tasks that power features like semantic search, recommendation systems, and retrieval-augmented generation (RAG) setups. Its performance benchmarks show it achieving high queries-per-second and fast index build times compared to similar tools. Because it runs in-process, developers can embed it in native apps, microservices, or edge computing scenarios where traditional server-based vector databases might be overkill. -
21
civitai
Open platform for sharing and discovering Stable Diffusion models
Civitai is an open source project that provides the codebase for a platform designed to share and manage generative AI models used for image generation. It focuses primarily on models compatible with Stable Diffusion and related technologies, allowing creators to upload, organize, and distribute custom AI models and related resources. These resources can include textual inversions, hypernetworks, aesthetic gradients, and variational autoencoders that modify or extend the capabilities of diffusion-based image generation systems. Civitai encourages collaboration by allowing users to publish their models, explore models created by others, and learn from the techniques used in the community. It also supports user accounts, model browsing, and interaction features that help creators showcase their work and receive feedback from other users. Developers can deploy the project to run their own instance of the platform and integrate with its available API to retrieve models. -
22
d2l-zh
Chinese-language edition of Dive into Deep Learning
d2l‑zh is the Chinese-language edition of Dive into Deep Learning, an interactive, open‑source deep learning textbook that combines code, math, and explanatory text. It features runnable Jupyter notebooks compatible with multiple frameworks (e.g., PyTorch, MXNet, TensorFlow), comprehensive theoretical analysis, and exercises. Widely adopted in over 70 countries and used by more than 500 universities for teaching deep learning. -
23
ort
Fast ML inference & training for ONNX models in Rust
ort is a high-performance Rust library that provides bindings to ONNX Runtime, enabling developers to run machine learning inference and training workflows directly within Rust applications using the standardized ONNX model format. It is designed to bridge the gap between modern machine learning frameworks and systems programming by offering a safe, ergonomic API for executing models originally built in ecosystems like PyTorch, TensorFlow, or scikit-learn. The library emphasizes speed and efficiency, leveraging hardware acceleration across CPUs, GPUs, and specialized accelerators to deliver low-latency inference both on-device and in server environments. One of its key strengths is its flexibility, as it supports multiple backends and allows developers to configure execution providers depending on available hardware. ort also includes advanced capabilities such as model compilation and optimization, reducing startup time and improving runtime performance in production systems. -
24
shimmy
Python-free Rust inference server
The shimmy project is a lightweight local inference server designed to run large language models with minimal overhead. Written primarily in Rust, the tool provides a small standalone binary that exposes an API compatible with the OpenAI interface, allowing existing applications to interact with local models without significant code changes. This compatibility enables developers to replace remote AI services with locally hosted models while keeping their existing software architecture intact. Shimmy focuses on performance and simplicity, using efficient runtime components to minimize memory usage and startup time compared to heavier inference frameworks. It supports modern model formats such as GGUF and SafeTensors and can automatically discover models stored locally or in common directories used by other AI tools. Advanced capabilities include CPU offloading for Mixture-of-Experts models and GPU acceleration, enabling large models to run on consumer hardware with limited VRAM. -
25
Grok-1
Open-source, high-performance Mixture-of-Experts large language model
Grok-1 is a 314-billion-parameter Mixture-of-Experts (MoE) large language model developed by xAI. Designed to optimize computational efficiency, it activates only 25% of its weights for each input token. In March 2024, xAI released Grok-1's model weights and architecture under the Apache 2.0 license, making them openly accessible to developers. The accompanying GitHub repository provides JAX example code for loading and running the model. Due to its substantial size, utilizing Grok-1 requires a machine with significant GPU memory. The repository's MoE layer implementation prioritizes correctness over efficiency, avoiding the need for custom kernels. This is a full repo snapshot ZIP file of the Grok-1 code.
