Open Source Python AI Models - Page 11
Sort By:
Python AI Models
View 730 business solutionsBrowse free open source Python AI Models and projects below. Use the toggles on the left to filter open source Python AI Models by OS, license, language, programming language, and project status.
-
Our Free Plans just got better! | Auth0You asked, we delivered! Auth0 is excited to expand our Free and Paid plans to include more options so you can focus on building, deploying, and scaling applications without having to worry about your security. Auth0 now, thank yourself later.
-
Custom VMs From 1 to 96 vCPUs With 99.95% UptimeLive migration and automatic failover keep workloads online through maintenance. One free e2-micro VM every month.
-
1
Step-Audio 2
Multi-modal large language model designed for audio understanding
Step-Audio2 is an advanced, end-to-end multimodal large language model designed for high-fidelity audio understanding and natural speech conversation: unlike many pipelines that separate speech recognition, processing, and synthesis, Step-Audio2 processes raw audio, reasons about semantic and paralinguistic content (like emotion, speaker characteristics, non-verbal cues), and can generate contextually appropriate responses — including potentially generating or transforming audio output. It integrates a latent-space audio encoder, discrete acoustic tokens, and reinforcement-learning–based training (CoT + RL) to enhance its ability to capture and reproduce voice styles, intonations, and subtle vocal cues. Moreover, Step-Audio2 supports tool-calling and retrieval-augmented generation (RAG), allowing it to access external knowledge sources or audio/text databases, thus reducing hallucinations and improving coherence in complex dialogues. -
2
Step-Audio-EditX
LLM-based Reinforcement Learning audio edit model
Step-Audio-EditX is an open-source, 3 billion-parameter audio model from StepFun AI designed to make expressive and precise editing of speech and audio as easy as text editing. Rather than treating audio editing as low-level waveform manipulation, this model converts speech into a sequence of discrete “audio tokens” (via a dual-codebook tokenizer) — combining a linguistic token stream and a semantic (prosody/emotion/style) token stream — thereby abstracting audio editing into high-level token operations. This allows users to modify not only what is said (the text) but also how it's said: emotion, tone, speaking style, prosody, accent, even paralinguistic cues. Because the model is trained with a “large-margin learning” objective over many synthesized and natural speech samples, it gains robust control over expressive attributes, and can perform iterative editing: e.g. you could record a line, then ask the model to “make it sadder,” “speak slower,” or “change accent to X.” -
3
Step-Video-T2V
State-of-the-art (SoTA) text-to-video pre-trained model
Step-Video-T2V is a state-of-the-art text-to-video foundation model developed to generate videos from natural-language prompts; its 30B-parameter architecture is designed to produce coherent, temporally extended video sequences — up to around 204 frames — based on input text. Under the hood it uses a compressed latent representation (a Video-VAE) to reduce spatial and temporal redundancy, and a denoising diffusion (or similar) process over that latent space to generate smooth, plausible motion and visuals. The model handles bilingual input (e.g. English and Chinese) thanks to dual encoders, and supports end-to-end text-to-video generation without requiring external assets. Its training and generation pipeline includes techniques like flow-matching, full 3D attention for temporal consistency, and fine-tuning approaches (e.g. video-based DPO) to improve fidelity and reduce artifacts. As a result, Step-Video-T2V aims to push the frontier of open-source video generation. -
4
Step1X-3D
High-Fidelity and Controllable Generation of Textured 3D Assets
Step1X-3D is an open-source framework for generating high-fidelity textured 3D assets from scratch — both their geometry and surface textures — using modern generative AI techniques. It combines a hybrid architecture: a geometry generation stage using a VAE-DiT model to output a watertight 3D representation (e.g. TSDF surface), and a texture synthesis stage that conditions on geometry and optionally reference input (or prompts) to produce view-consistent textures using a diffusion-based texture module. The result is fully 3D assets — meshes + textures — which can be rendered from any viewpoint, textured consistently, and used in 3D applications. To achieve this, the project includes a massive curated dataset: among more than 5 million candidate 3D assets, it filters and standardizes to produce a high-quality 2 million–asset subset suitable for training. -
Build Agents and Models on One PlatformGemini Enterprise Agent Platform is Google Cloud's comprehensive platform for developers to build, scale, govern, and optimize agents and models. Choose from Google's most advanced models and third-party models like Anthropic's Claude Model Family.
-
5
Step1X-Edit
A SOTA open-source image editing model
Step1X-Edit is a state-of-the-art open-source image editing model/framework that uses a multimodal large language model (LLM) together with a diffusion-based image decoder to let users edit images simply via natural-language instructions plus a reference image. You supply an existing image and a textual command — e.g. “add a ruby pendant on the girl’s neck” or “make the background a sunset over mountains” — and the model interprets the instruction, computes a latent embedding combining the image content and user intent, then decodes a new image implementing the edit. The model targets general-purpose editing: from object addition/removal, style changes, recoloring, retouching, background replacement, to complex transformations like changing lighting, mood, or art style. The authors trained it on a large curated dataset and benchmarked it on a newly introduced evaluation suite, showing that Step1X-Edit significantly outperforms previous open-source baselines. -
6
Style Aligned
Official code for Style Aligned Image Generation via Shared Attention
StyleAligned is a diffusion-model editing technique and codebase that preserves the visual “style” of an original image while applying new semantic edits driven by text. Instead of fully re-generating an image—and risking changes to lighting, texture, or rendering choices—the method aligns internal features across denoising steps so the target edit inherits the source style. This alignment acts like a constraint on the model’s evolution, steering composition, palette, and brushwork even as objects or attributes change. The result is more consistent edits across a set, which is crucial for workflows like product variations, character sheets, or brand-coherent art. The repository provides reproducible scripts, reference prompts, and guidance for tuning strengths so users can dial in subtle retouches or bolder substitutions. Because it builds on widely used diffusion checkpoints, creators can integrate it without training or dataset collection. -
7
Surya
Implementation of the Surya Foundation Model for Heliophysics
Surya is an open‑source, AI‑based foundation model for heliophysics developed collaboratively by NASA (via the IMPACT AI team) and IBM. Named after the Sanskrit word for “sun,” Surya is trained on nine years of high‑resolution solar imagery from NASA’s Solar Dynamics Observatory (SDO). It is designed to forecast solar phenomena—such as flares, solar wind, irradiance, and active region behavior—by predicting future solar images with a sophisticated long–short vision transformer architecture, thereby enabling improved space weather forecasting. Foresees solar flares, wind, EUV spectra, and active region formation in advance. Achieves approximately 16% improvement in forecasting accuracy over traditional methods. 366-million‑parameter foundation model capturing general-purpose solar representations. -
8
TADA
Open Source Speech Language Model
TADA is an open-source speech-language modeling framework designed to unify spoken audio and text representations within a single generative architecture. The system focuses on aligning speech and text streams using a dual-alignment mechanism that synchronizes the acoustic signal with its textual representation. By modeling both modalities together, the framework allows developers to build systems capable of generating, understanding, and transforming speech and language simultaneously. This approach can support applications such as conversational AI, speech synthesis, multimodal language modeling, and speech understanding systems. The project explores ways to treat speech and text as integrated data streams rather than separate pipelines, enabling more coherent interactions between language and audio. Because it operates as a generative framework, TADA can be used for research into advanced speech-language models and multimodal artificial intelligence systems. -
9
Tiktoken
tiktoken is a fast BPE tokeniser for use with OpenAI's models
tiktoken is a high-performance, tokenizer library (based on byte-pair encoding, BPE) designed for use with OpenAI’s models. It handles encoding and decoding text to token IDs efficiently, with minimal overhead. Because tokenization is a fundamental step in preparing text for models, tiktoken is optimized for speed, memory, and correctness in model contexts (e.g. matching OpenAI’s internal tokenization). The repo supports multiple encodings (e.g. “cl100k_base”) and lets users switch encoding names to match different model contexts. It also offers extension mechanisms so that custom encodings can be registered. Internally, it includes the core tokenizer logic (often implemented in Rust or efficient lower-level code), APIs for encoding, decoding, and counting tokens, and binding layers to Python (and sometimes other languages) for easy use. -
99.99% Uptime for MySQL and PostgreSQL DatabasesCloud SQL Enterprise Plus delivers near-zero downtime with 35 days of point-in-time recovery. Supports MySQL, PostgreSQL, and SQL Server.
-
10
TimeSformer
The official pytorch implementation of our paper
TimeSformer is a vision transformer architecture for video that extends the standard attention mechanism into spatiotemporal attention. The model alternates attention along spatial and temporal dimensions (or designs variants like divided attention) so that it can capture both appearance and motion cues in video. Because the attention is global across frames, TimeSformer can reason about dependencies across long time spans, not just local neighborhoods. The official implementation in PyTorch provides configurations, pretrained models, and training scripts that make it straightforward to evaluate or fine-tune on video datasets. TimeSformer was influential in showing that pure transformer architectures—without convolutional backbones—can perform strongly on video classification tasks. Its flexible attention design allows experimenting with different factoring (spatial-then-temporal, joint, etc.) to trade off compute, memory, and accuracy. -
11
TimesFM
Pretrained time-series foundation model developed by Google Research
TimesFM is a pretrained time-series foundation model from Google Research built for forecasting tasks, designed to generalize across many domains without requiring extensive per-dataset retraining. It provides a decoder-only model approach to forecasting, aiming for strong performance even in zero-shot or low-data settings where traditional models often struggle. The project includes code and an inference API intended to make it practical to run forecasts programmatically, with options to use different backends such as Torch or Flax depending on your environment and performance needs. Newer releases emphasize expanded context handling and more flexible forecasting outputs, including quantile forecasting so users can get uncertainty estimates rather than only point predictions. The repository also documents how model versions evolved, with newer variants focusing on efficiency and longer context windows while maintaining forecasting quality. -
12
ToMe (Token Merging)
A method to increase the speed and lower the memory footprint
ToMe (Token Merging) is a PyTorch-based optimization framework designed to significantly accelerate Vision Transformer (ViT) architectures without retraining. Developed by researchers at Facebook (Meta AI), ToMe introduces an efficient technique that merges similar tokens within transformer layers, reducing redundant computation while preserving model accuracy. This approach differs from token pruning, which removes background tokens entirely; instead, ToMe merges tokens based on feature similarity, allowing it to compress both foreground and background information efficiently. ToMe integrates seamlessly into existing transformer models such as DeiT, MAE, SWAG, and timm ViTs, offering 2–3x speedups during inference and substantial efficiency gains during training. The method can be applied dynamically at inference time or incorporated into training for improved performance. -
13
Tracking Any Point (TAP)
DeepMind model for tracking arbitrary points across videos & robotics
TAPNet is the official Google DeepMind repository for Tracking Any Point (TAP), bundling datasets, models, benchmarks, and demos for precise point tracking in videos. The project includes the TAP-Vid and TAPVid-3D benchmarks, which evaluate long-range tracking of arbitrary points in 2D and 3D across diverse real and synthetic videos. Its flagship models—TAPIR, BootsTAPIR, and the latest TAPNext—use matching plus temporal refinement or next-token style propagation to achieve state-of-the-art accuracy and speed on TAP-Vid. RoboTAP demonstrates how TAPIR-style tracks can drive real-world robot manipulation via efficient imitation, and ships with a dataset of annotated robotics videos. The repo provides JAX and PyTorch checkpoints, Colab demos, and a real-time live demo that runs on a GPU to let you select and track points interactively. -
14
Transformer Debugger
Tool for exploring and debugging transformer model behaviors
Transformer Debugger (TDB) is a research tool developed by OpenAI’s Superalignment team to investigate and interpret the behaviors of small language models. It combines automated interpretability methods with sparse autoencoders, enabling researchers to analyze how specific neurons, attention heads, and latent features contribute to a model’s outputs. TDB allows users to intervene directly in the forward pass of a model and observe how such interventions change predictions, making it possible to answer questions like why a token was selected or why an attention head focused on a certain input. It automatically identifies and explains the most influential components, highlights activation patterns, and maps relationships across circuits within the model. The tool includes both a React-based neuron viewer for exploring model components and a backend activation server for running inferences and serving data. -
15
UCO3D
Uncommon Objects in 3D dataset
uCO3D is a large-scale 3D vision dataset and toolkit centered on turn-table videos of everyday objects drawn from the LVIS taxonomy. It provides about 170,000 full videos per object instance rather than still frames, along with per-video annotations including object masks, calibrated camera poses, and multiple flavors of point clouds. Each sequence also ships with a precomputed 3D Gaussian Splat reconstruction, enabling fast, differentiable rendering workflows and modern implicit/point-based modeling experiments. The repository includes automated downloaders with checksum verification, fine-grained controls to fetch only selected modalities or super-categories, and a lightweight Python API for loading frames, geometry, and splats on demand. Metadata is indexed in SQLite for quick queries at scale, and helper builders handle alignment, undistortion, frame extraction from videos, and cropping around the object. -
16
VALL-E
PyTorch implementation of VALL-E (Zero-Shot Text-To-Speech)
We introduce a language modeling approach for text to speech synthesis (TTS). Specifically, we train a neural codec language model (called VALL-E) using discrete codes derived from an off-the-shelf neural audio codec model, and regard TTS as a conditional language modeling task rather than continuous signal regression as in previous work. During the pre-training stage, we scale up the TTS training data to 60K hours of English speech which is hundreds of times larger than existing systems. VALL-E emerges in-context learning capabilities and can be used to synthesize high-quality personalized speech with only a 3-second enrolled recording of an unseen speaker as an acoustic prompt. Experiment results show that VALL-E significantly outperforms the state-of-the-art zero-shot TTS system in terms of speech naturalness and speaker similarity. In addition, we find VALL-E could preserve the speaker's emotion and acoustic environment of the acoustic prompt in synthesis. -
17
VGGSfM
VGGSfM: Visual Geometry Grounded Deep Structure From Motion
VGGSfM is an advanced structure-from-motion (SfM) framework jointly developed by Meta AI Research (GenAI) and the University of Oxford’s Visual Geometry Group (VGG). It reconstructs 3D geometry, dense depth, and camera poses directly from unordered or sequential images and videos. The system combines learned feature matching and geometric optimization to generate high-quality camera calibrations, sparse/dense point clouds, and depth maps in standard COLMAP format. Version 2.0 adds support for dynamic scene handling, dense point cloud export, video-based reconstruction (1000+ frames), and integration with Gaussian Splatting pipelines. It leverages tools like PyCOLMAP, poselib, LightGlue, and PyTorch3D for feature matching, pose estimation, and visualization. With minimal configuration, users can process single scenes or full video sequences, apply motion masks to exclude moving objects, and train neural radiance or splatting models directly from reconstructed outputs. -
18
VibeThinker
Diversity-driven optimization and large-model reasoning ability
VibeThinker is a compact but high-capability open-source language model released by WeiboAI (Sina AI Lab). It contains about 1.5 billion parameters, far smaller than many “frontier” models, yet it is explicitly optimized for reasoning, mathematics, and code generation tasks rather than general open-domain chat. The innovation lies in its training methodology: the team uses what they call the Spectrum-to-Signal Principle (SSP), where a first stage emphasizes diversity of reasoning paths (the “spectrum” phase) and a second stage uses reinforcement techniques (the “signal” phase) to refine toward correctness and strong reasoning. The result is a model that outpaces many much larger models on domain-specific benchmarks, demonstrating that smaller models, if trained carefully and with the right objectives, can achieve high performance in reasoning-centric tasks. -
19
Video Pre-Training
Learning to Act by Watching Unlabeled Online Videos
The Video PreTraining (VPT) repository provides code and model artifacts for a project where agents learn to act by watching human gameplay videos—specifically, gameplay of Minecraft—using behavioral cloning. The idea is to learn general priors of control from large-scale, unlabeled video data, and then optionally fine-tune those priors for more goal-directed behavior via environment interaction. The repository contains demonstration models of different widths, fine-tuned variants (e.g. for building houses or early-game tasks), and inference scripts that instantiate agents from pretrained weights. Key modules include the behavioral cloning logic, the agent wrapper, and data loading pipelines (with an accessible skeleton for loading Minecraft demonstration data). The repo also includes a run_agent.py script for testing an agent interactively, and an agent.py module encapsulating the control logic. -
20
Vidi2
Large Multimodal Models for Video Understanding and Editing
Vidi is a family of large multimodal models developed for deep video understanding and editing tasks, integrating vision, audio, and language to allow sophisticated querying and manipulation of video content. It’s designed to process long-form, real-world videos and answer complex queries such as “when in this clip does X happen?” or “where in the frame is object Y during that moment?” — offering temporal retrieval, spatio-temporal grounding (i.e. locating objects over time + space), and even video question answering. Vidi targets applications like intelligent video editing, automated video search, content analysis, and editing assistance, enabling users to efficiently locate relevant segments and objects in hours-long footage. The system is built with open-source release in mind, giving developers access to model code, inference scripts, and evaluation pipelines so they can reproduce research results or integrate Vidi into their own video-processing workflows. -
21
VisualGLM-6B
Chinese and English multimodal conversational language model
VisualGLM-6B is an open-source multimodal conversational language model developed by ZhipuAI that supports both images and text in Chinese and English. It builds on the ChatGLM-6B backbone, with 6.2 billion language parameters, and incorporates a BLIP2-Qformer visual module to connect vision and language. In total, the model has 7.8 billion parameters. Trained on a large bilingual dataset — including 30 million high-quality Chinese image-text pairs from CogView and 300 million English pairs — VisualGLM-6B is designed for image understanding, description, and question answering. Fine-tuning on long visual QA datasets further aligns the model’s responses with human preferences. The repository provides inference APIs, command-line demos, web demos, and efficient fine-tuning options like LoRA, QLoRA, and P-tuning. It also supports quantization down to INT4, enabling local deployment on consumer GPUs with as little as 6.3 GB VRAM. -
22
Watermark Anything
Official implementation of Watermark Anything with Localized Messages
Watermark Anything (WAM) is an advanced deep learning framework for embedding and detecting localized watermarks in digital images. Developed by Facebook Research, it provides a robust, flexible system that allows users to insert one or multiple watermarks within selected image regions while maintaining visual quality and recoverability. Unlike traditional watermarking methods that rely on uniform embedding, WAM supports spatially localized watermarks, enabling targeted protection of specific image regions or objects. The model is trained to balance imperceptibility, ensuring minimal visual distortion, with robustness against transformations and edits such as cropping or motion. -
23
YOLOv4
PyTorch implementation of YOLOv4
PyTorch_YOLOv4 is a PyTorch implementation of YOLOv4 based on the earlier ultralytics YOLOv3 codebase. It provides a practical way to train, test, and run YOLOv4-style object detection models without relying only on the original Darknet implementation. The repository supports common detection workflows such as dataset preparation, model training, evaluation, inference, and weight conversion. It is useful for developers who prefer the PyTorch ecosystem for experimentation, debugging, and integration with other machine learning tooling. The project also connects to the broader YOLOv4 family, including CSP-based architecture ideas and real-time detection improvements. It is best suited for researchers and engineers who want YOLOv4 behavior in a Python-first deep learning environment. -
24
Z80-μLM
Z80-μLM is a 2-bit quantized language model
Z80-μLM is a retro-computing AI project that demonstrates a tiny language model (Z80-μLM) engineered to run on an 8-bit Z80 CPU by aggressively quantizing weights down to 2-bit precision. The repository provides a complete workflow where you train or fine-tune conversational models in Python, then export them into a format that can be executed on classic Z80 systems. A key deliverable is producing CP/M-compatible .COM binaries, enabling a genuinely vintage “chat with your computer” experience on real hardware or accurate emulators. The project sits at the intersection of machine learning and systems constraints, showing how model architecture, quantization, and inference code generation can be adapted to extreme memory and compute limits. It also functions as an educational reference for how to reduce inference to operations that fit an old-school instruction set and runtime environment. -
25
fairseq2
FAIR Sequence Modeling Toolkit 2
fairseq2 is a modern, modular sequence modeling framework developed by Meta AI Research as a complete redesign of the original fairseq library. Built from the ground up for scalability, composability, and research flexibility, fairseq2 supports a broad range of language, speech, and multimodal content generation tasks, including instruction fine-tuning, reinforcement learning from human feedback (RLHF), and large-scale multilingual modeling. Unlike the original fairseq—which evolved into a large, monolithic codebase—fairseq2 introduces a clean, plugin-oriented architecture designed for long-term maintainability and rapid experimentation. It supports multi-GPU and multi-node distributed training using DDP, FSDP, and tensor parallelism, capable of scaling up to 70B+ parameter models. The framework integrates seamlessly with PyTorch 2.x features such as torch.compile, Fully Sharded Data Parallel (FSDP), and modern configuration management.
