StarBoard

Moving right now, since the previous snapshot

#RepoCategory LanguageAge StarsNow24h7d
870 VectifyAI/PageIndex ↗
📑 PageIndex: Document Index for Vectorless, Reasoning-based RAG
 LLM Models, Inference & Training Python 545d 36543 - +652 +749
1002 NandhaKishorM/laya ↗
Non-autoregressive System 1 decision engine. Typed choice, score and yes/no decisions over any text in a single forward pass, in 100+ languages, with a router that picks the right checkpoint per request.
 LLM Models, Inference & Training Python 11d 27941 - +875 +18337
1035 jaredpalmer/kev ↗
Jev-like family of decision models built on top of Qwen3.5/3.8 you can train and run on your own
 LLM Models, Inference & Training Python 11d 7713 - +208 +5416
81 ggml-org/llama.cpp ↗
LLM inference in C/C++
 LLM Models, Inference & Training C++ 1298d 129830 - +87 +741
339 jingyaogong/minimind ↗
🧠 Train a 64M-parameter LLM from scratch in just 2h!
 LLM Models, Inference & Training Python 794d 62862 - +67 +840
50 huggingface/transformers ↗
🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.
LLM Models, Inference & Training Python 2891d 166790 - +42 +313
647 Wei-Shaw/sub2api ↗
Sub2API 一站式开源中转服务，让 Claude、Openai 、Gemini、Grok订阅统一接入，支持拼车共享，更高效分摊成本，原生工具无缝使用。
 LLM Models, Inference & Training Go 285d 43024 - +74 +748
1165 YuJunZhiXue/dsh-purge ↗
DeepSeek Harness 破甲：让所有模型都能破甲，不同模型可换不同提示词；默认提示词面向国模「小码酱」。Jailbreak for every model — swap prompts per model. 求 Star 收藏 ⭐
 LLM Models, Inference & Training JavaScript 40d 2591 - +95 +825
152 vllm-project/vllm ↗
A high-throughput and memory-efficient inference and serving engine for LLMs
 LLM Models, Inference & Training Python 1327d 92912 - +71 +555
520 QuantumNous/new-api ↗
A unified AI model hub for aggregation & distribution. It supports cross-converting various LLMs into OpenAI-compatible, Claude-compatible, or Gemini-compatible formats. A centralized gateway for personal and enterprise model management.
 LLM Models, Inference & Training Go 1053d 49040 - +42 +432
223 unslothai/unsloth ↗
Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4, MiniMax-H3, Gemma 4, FLUX and more.
 LLM Models, Inference & Training Python 1034d 76990 - +98 +450
798 JustVugg/colibri ↗
Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦
 LLM Models, Inference & Training C 89d 38193 - +170 +1373
43 ollama/ollama ↗
Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
 LLM Models, Inference & Training Go 1190d 181894 - +58 +501
241 hiyouga/LlamaFactory ↗
Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)
 LLM Models, Inference & Training Python 1219d 75169 - +37 +211
376 BerriAI/litellm ↗
The fastest, litest AI Gateway. Rust core with Python SDK. Call 100+ LLM APIs in OpenAI (or native) format with cost tracking, guardrails, load balancing, and logging [Bedrock, Azure, OpenAI, Anthropic, OpenAI, VertexAI, vLLM, Nvidia NIM]
 LLM Models, Inference & Training Python 1160d 59825 - +67 +483
735 HKUDS/LightRAG ↗
[EMNLP2025] LightRAG: Simple and Fast Retrieval-Augmented Generation
 LLM Models, Inference & Training Python 726d 39913 - +18 +110
747 shiyu-coder/Kronos ↗
Kronos: A Foundation Model for the Language of Financial Markets
 LLM Models, Inference & Training Python 455d 39627 - +99 +308
835 AlexsJones/llmfit ↗
Hundreds of models & providers. One command to find what runs on your hardware.
 LLM Models, Inference & Training Rust 225d 37287 - +63 +345
926 lyogavin/airllm ↗
AirLLM 70B inference with single 4GB GPU
 LLM Models, Inference & Training Jupyter Notebook 1204d 35179 - +44 +532
986 google-research/timesfm ↗
TimesFM (Time Series Foundation Model) is a pretrained time-series foundation model developed by Google Research for time-series forecasting.
 LLM Models, Inference & Training Python 882d 33953 - +61 -
1048 mizorewww/laya-mlx ↗
Native MLX runtime for Laya typed decision models — 7–14 ms short decisions on M3 Max. No text generation, PyTorch, or cloud API.
 LLM Models, Inference & Training Python 9d 6593 - +78 +3018
29 tensorflow/tensorflow ↗
An Open Source Machine Learning Framework for Everyone
 LLM Models, Inference & Training C++ 3979d 200603 - +26 +373
51 AUTOMATIC1111/stable-diffusion-webui ↗
Stable Diffusion web UI
 LLM Models, Inference & Training Python 1498d 165147 - 0 +102
121 pytorch/pytorch ↗
Tensors and Dynamic neural networks in Python with strong GPU acceleration
 LLM Models, Inference & Training Python 3699d 103485 - +51 +330
391 karpathy/nanochat ↗
The best ChatGPT that $100 can buy.
 LLM Models, Inference & Training Python 350d 58318 - +22 +115
738 2noise/ChatTTS ↗
A generative speech model for daily dialogue.
 LLM Models, Inference & Training Python 854d 39880 - +4 +19
793 stanfordnlp/dspy ↗
DSPy: The framework for programming—not prompting—language models
 LLM Models, Inference & Training Python 1358d 38410 - +21 +226
989 SillyTavern/SillyTavern ↗
LLM Frontend for Power Users.
 LLM Models, Inference & Training JavaScript 1327d 33900 - +41 +257
1026 MiniMax-AI/MiniMax-H3 ↗
LLM Models, Inference & Training Python 60d 9350 - +41 +305
109 openai/whisper ↗
Robust Speech Recognition via Large-Scale Weak Supervision
 LLM Models, Inference & Training Python 1473d 109714 - +39 +273
258 CompVis/stable-diffusion ↗
A latent text-to-image diffusion model
 LLM Models, Inference & Training Jupyter Notebook 1510d 73489 - +4 +34
323 keras-team/keras ↗
Deep Learning for humans
 LLM Models, Inference & Training Python 4203d 64343 - -2 +17
441 ggml-org/whisper.cpp ↗
Port of OpenAI's Whisper model in C/C++
 LLM Models, Inference & Training C++ 1464d 53999 - +22 +157
453 lllyasviel/Fooocus ↗
Focus on prompting and generating
 LLM Models, Inference & Training Python 1146d 53224 - +19 +88
548 exo-explore/exo ↗
Run frontier AI locally.
 LLM Models, Inference & Training Python 826d 47673 - +10 +79
631 ray-project/ray ↗
Ray is an AI compute engine. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.
 LLM Models, Inference & Training Python 3625d 43940 - +1 +56
641 chatanywhere/GPT_API_free ↗
免费大模型API，支持免费调用GPT、DeepSeek等主流模型，免费额度10000点，每日刷新！另付费价格最低官方1-2折！
 LLM Models, Inference & Training 1254d 43342 - +51 +354
783 RVC-Project/Retrieval-based-Voice-Conversion-WebUI ↗
Easily train a good VC model with voice data <= 10 mins!
 LLM Models, Inference & Training Python 1281d 38615 - +21 +191
804 OpenBMB/VoxCPM ↗
VoxCPM2: Tokenizer-Free TTS for Multilingual Speech Generation, Creative Voice Design, and True-to-Life Cloning
 LLM Models, Inference & Training Python 378d 38119 - +90 +265
815 myshell-ai/OpenVoice ↗
Instant voice cloning by MIT and MyShell. Audio foundation model.
 LLM Models, Inference & Training Python 1034d 37717 - +17 +111
879 jax-ml/jax ↗
Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more
 LLM Models, Inference & Training Python 2895d 36362 - +5 +42
1014 FlashML-org/FreeToken ↗
FreeToken brings datacenter-scale model serving to your desktop. Run massive models locally, fast and efficiently.
 LLM Models, Inference & Training Python 70d 13955 - +36 +505
1029 MoonshotAI/Kimi-K3 ↗
Open Frontier Intelligence
 LLM Models, Inference & Training 63d 8876 - +5 +40
1030 FareedKhan-dev/kimi-k3-in-c ↗
A 2.78-trillion-parameter Kimi K3 running inference on a single CPU in 8.24 GB of RAM. Portable C99: no BLAS, no framework, no GPU.
 LLM Models, Inference & Training C 58d 8781 - +44 +647
1087 yynxxxxx/Codex-X ↗
OpenAI Codex 桌面端/CLI 的可视化管理工具，具有Provider/API 切换、会话同步、提示词注入、Skills/MCP 管理、TOML 配置可视化的跨平台工具。
 LLM Models, Inference & Training Rust 87d 3988 - +21 +337
1240 yibie/awesome-jev ↗
A curated list of public projects, integrations, and discussions built on Jev — TypeSafe AI's System One model for typed decisions.
 LLM Models, Inference & Training Python 11d 1941 - +69 -
1278 Tencent/WeMM-Embedding ↗
WeMM-Embedding is a family of universal multimodal embedding models by the WeChat Vision Team at Tencent, supporting multimodal understanding and retrieval.
 LLM Models, Inference & Training Python 35d 1689 - +4 +45

LLM Models, Inference & Training, movers in the last 24 hours

#RepoCategory LanguageAge StarsNow24h7d
1002 NandhaKishorM/laya ↗
Non-autoregressive System 1 decision engine. Typed choice, score and yes/no decisions over any text in a single forward pass, in 100+ languages, with a router that picks the right checkpoint per request.
 LLM Models, Inference & Training Python 11d 27941 - +875 +18337
870 VectifyAI/PageIndex ↗
📑 PageIndex: Document Index for Vectorless, Reasoning-based RAG
 LLM Models, Inference & Training Python 545d 36543 - +652 +749
1035 jaredpalmer/kev ↗
Jev-like family of decision models built on top of Qwen3.5/3.8 you can train and run on your own
 LLM Models, Inference & Training Python 11d 7713 - +208 +5416
798 JustVugg/colibri ↗
Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦
 LLM Models, Inference & Training C 89d 38193 - +170 +1373
747 shiyu-coder/Kronos ↗
Kronos: A Foundation Model for the Language of Financial Markets
 LLM Models, Inference & Training Python 455d 39627 - +99 +308
223 unslothai/unsloth ↗
Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4, MiniMax-H3, Gemma 4, FLUX and more.
 LLM Models, Inference & Training Python 1034d 76990 - +98 +450
1165 YuJunZhiXue/dsh-purge ↗
DeepSeek Harness 破甲：让所有模型都能破甲，不同模型可换不同提示词；默认提示词面向国模「小码酱」。Jailbreak for every model — swap prompts per model. 求 Star 收藏 ⭐
 LLM Models, Inference & Training JavaScript 40d 2591 - +95 +825
804 OpenBMB/VoxCPM ↗
VoxCPM2: Tokenizer-Free TTS for Multilingual Speech Generation, Creative Voice Design, and True-to-Life Cloning
 LLM Models, Inference & Training Python 378d 38119 - +90 +265
81 ggml-org/llama.cpp ↗
LLM inference in C/C++
 LLM Models, Inference & Training C++ 1298d 129830 - +87 +741
1048 mizorewww/laya-mlx ↗
Native MLX runtime for Laya typed decision models — 7–14 ms short decisions on M3 Max. No text generation, PyTorch, or cloud API.
 LLM Models, Inference & Training Python 9d 6593 - +78 +3018
647 Wei-Shaw/sub2api ↗
Sub2API 一站式开源中转服务，让 Claude、Openai 、Gemini、Grok订阅统一接入，支持拼车共享，更高效分摊成本，原生工具无缝使用。
 LLM Models, Inference & Training Go 285d 43024 - +74 +748
152 vllm-project/vllm ↗
A high-throughput and memory-efficient inference and serving engine for LLMs
 LLM Models, Inference & Training Python 1327d 92912 - +71 +555
1240 yibie/awesome-jev ↗
A curated list of public projects, integrations, and discussions built on Jev — TypeSafe AI's System One model for typed decisions.
 LLM Models, Inference & Training Python 11d 1941 - +69 -
339 jingyaogong/minimind ↗
🧠 Train a 64M-parameter LLM from scratch in just 2h!
 LLM Models, Inference & Training Python 794d 62862 - +67 +840
376 BerriAI/litellm ↗
The fastest, litest AI Gateway. Rust core with Python SDK. Call 100+ LLM APIs in OpenAI (or native) format with cost tracking, guardrails, load balancing, and logging [Bedrock, Azure, OpenAI, Anthropic, OpenAI, VertexAI, vLLM, Nvidia NIM]
 LLM Models, Inference & Training Python 1160d 59825 - +67 +483
835 AlexsJones/llmfit ↗
Hundreds of models & providers. One command to find what runs on your hardware.
 LLM Models, Inference & Training Rust 225d 37287 - +63 +345
986 google-research/timesfm ↗
TimesFM (Time Series Foundation Model) is a pretrained time-series foundation model developed by Google Research for time-series forecasting.
 LLM Models, Inference & Training Python 882d 33953 - +61 -
43 ollama/ollama ↗
Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
 LLM Models, Inference & Training Go 1190d 181894 - +58 +501
869 sgl-project/sglang ↗
SGLang is a high-performance serving framework for large language models and multimodal models.
 LLM Models, Inference & Training Python 995d 36558 - +52 +287
121 pytorch/pytorch ↗
Tensors and Dynamic neural networks in Python with strong GPU acceleration
 LLM Models, Inference & Training Python 3699d 103485 - +51 +330
641 chatanywhere/GPT_API_free ↗
免费大模型API，支持免费调用GPT、DeepSeek等主流模型，免费额度10000点，每日刷新！另付费价格最低官方1-2折！
 LLM Models, Inference & Training 1254d 43342 - +51 +354
1239 Taichu-AI/ZDTaichu5.0-9B ↗
LLM Models, Inference & Training Python 25d 1942 - +45 -
926 lyogavin/airllm ↗
AirLLM 70B inference with single 4GB GPU
 LLM Models, Inference & Training Jupyter Notebook 1204d 35179 - +44 +532
1030 FareedKhan-dev/kimi-k3-in-c ↗
A 2.78-trillion-parameter Kimi K3 running inference on a single CPU in 8.24 GB of RAM. Portable C99: no BLAS, no framework, no GPU.
 LLM Models, Inference & Training C 58d 8781 - +44 +647
50 huggingface/transformers ↗
🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.
LLM Models, Inference & Training Python 2891d 166790 - +42 +313
520 QuantumNous/new-api ↗
A unified AI model hub for aggregation & distribution. It supports cross-converting various LLMs into OpenAI-compatible, Claude-compatible, or Gemini-compatible formats. A centralized gateway for personal and enterprise model management.
 LLM Models, Inference & Training Go 1053d 49040 - +42 +432
989 SillyTavern/SillyTavern ↗
LLM Frontend for Power Users.
 LLM Models, Inference & Training JavaScript 1327d 33900 - +41 +257
1026 MiniMax-AI/MiniMax-H3 ↗
LLM Models, Inference & Training Python 60d 9350 - +41 +305
109 openai/whisper ↗
Robust Speech Recognition via Large-Scale Weak Supervision
 LLM Models, Inference & Training Python 1473d 109714 - +39 +273
241 hiyouga/LlamaFactory ↗
Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)
 LLM Models, Inference & Training Python 1219d 75169 - +37 +211
1014 FlashML-org/FreeToken ↗
FreeToken brings datacenter-scale model serving to your desktop. Run massive models locally, fast and efficiently.
 LLM Models, Inference & Training Python 70d 13955 - +36 +505
1291 QwenLM/Qwen-Image-2.1 ↗
Qwen's most powerful open-source image generation model
 LLM Models, Inference & Training Python 15d 1609 - +27 -
29 tensorflow/tensorflow ↗
An Open Source Machine Learning Framework for Everyone
 LLM Models, Inference & Training C++ 3979d 200603 - +26 +373
391 karpathy/nanochat ↗
The best ChatGPT that $100 can buy.
 LLM Models, Inference & Training Python 350d 58318 - +22 +115
441 ggml-org/whisper.cpp ↗
Port of OpenAI's Whisper model in C/C++
 LLM Models, Inference & Training C++ 1464d 53999 - +22 +157
783 RVC-Project/Retrieval-based-Voice-Conversion-WebUI ↗
Easily train a good VC model with voice data <= 10 mins!
 LLM Models, Inference & Training Python 1281d 38615 - +21 +191
793 stanfordnlp/dspy ↗
DSPy: The framework for programming—not prompting—language models
 LLM Models, Inference & Training Python 1358d 38410 - +21 +226
1087 yynxxxxx/Codex-X ↗
OpenAI Codex 桌面端/CLI 的可视化管理工具，具有Provider/API 切换、会话同步、提示词注入、Skills/MCP 管理、TOML 配置可视化的跨平台工具。
 LLM Models, Inference & Training Rust 87d 3988 - +21 +337
453 lllyasviel/Fooocus ↗
Focus on prompting and generating
 LLM Models, Inference & Training Python 1146d 53224 - +19 +88
735 HKUDS/LightRAG ↗
[EMNLP2025] LightRAG: Simple and Fast Retrieval-Augmented Generation
 LLM Models, Inference & Training Python 726d 39913 - +18 +110
328 karpathy/nanoGPT ↗
The simplest, fastest repository for training/finetuning medium-sized GPTs.
 LLM Models, Inference & Training Python 1371d 63437 - +17 +144
815 myshell-ai/OpenVoice ↗
Instant voice cloning by MIT and MyShell. Audio foundation model.
 LLM Models, Inference & Training Python 1034d 37717 - +17 +111
345 RVC-Boss/GPT-SoVITS ↗
1 min voice data can also be used to train a good TTS model! (few shot voice cloning)
 LLM Models, Inference & Training Python 988d 62239 - +16 +223
611 janhq/jan ↗
Jan is an open source alternative to ChatGPT that runs 100% offline on your computer.
 LLM Models, Inference & Training Rust 1139d 44698 - +16 +113
887 microsoft/graphrag ↗
A modular graph-based Retrieval-Augmented Generation (RAG) system
 LLM Models, Inference & Training Python 915d 36149 - +16 +89
1269 syv-ai/HyperQwen ↗
Serve large Qwen models fast on the GPUs you actually own. Qwen3.8-27B on a single 24 GB card with vLLM: 127 tok/s single-user (381 when the answer quotes the prompt), ~1,035 tok/s at 64 concurrent, 150k-262k context. vLLM patches, requant pipeline, benchmarks.
 LLM Models, Inference & Training Python 44d 1755 - +16 +144
848 google-ai-edge/mediapipe ↗
Cross-platform, customizable ML solutions for live and streaming media.
 LLM Models, Inference & Training C++ 2664d 37114 - +14 +86
859 xinntao/Real-ESRGAN ↗
Real-ESRGAN aims at developing Practical Algorithms for General Image/Video Restoration.
 LLM Models, Inference & Training Python 1898d 36939 - +13 +72
509 GokuMohandas/Made-With-ML ↗
Learn how to develop, deploy and iterate on production-grade ML applications.
 LLM Models, Inference & Training Jupyter Notebook 2885d 49643 - +11 +70
514 mudler/LocalAI ↗
LocalAI is the open-source AI engine. Run any model - LLMs, vision, voice, image, video - on any hardware. No GPU required.
 LLM Models, Inference & Training Go 1290d 49315 - +11 +108

LLM Models, Inference & Training, movers in the last 7 days

#RepoCategory LanguageAge StarsNow24h7d
1002 NandhaKishorM/laya ↗
Non-autoregressive System 1 decision engine. Typed choice, score and yes/no decisions over any text in a single forward pass, in 100+ languages, with a router that picks the right checkpoint per request.
 LLM Models, Inference & Training Python 11d 27941 - +875 +18337
1035 jaredpalmer/kev ↗
Jev-like family of decision models built on top of Qwen3.5/3.8 you can train and run on your own
 LLM Models, Inference & Training Python 11d 7713 - +208 +5416
1048 mizorewww/laya-mlx ↗
Native MLX runtime for Laya typed decision models — 7–14 ms short decisions on M3 Max. No text generation, PyTorch, or cloud API.
 LLM Models, Inference & Training Python 9d 6593 - +78 +3018
798 JustVugg/colibri ↗
Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦
 LLM Models, Inference & Training C 89d 38193 - +170 +1373
339 jingyaogong/minimind ↗
🧠 Train a 64M-parameter LLM from scratch in just 2h!
 LLM Models, Inference & Training Python 794d 62862 - +67 +840
1165 YuJunZhiXue/dsh-purge ↗
DeepSeek Harness 破甲：让所有模型都能破甲，不同模型可换不同提示词；默认提示词面向国模「小码酱」。Jailbreak for every model — swap prompts per model. 求 Star 收藏 ⭐
 LLM Models, Inference & Training JavaScript 40d 2591 - +95 +825
870 VectifyAI/PageIndex ↗
📑 PageIndex: Document Index for Vectorless, Reasoning-based RAG
 LLM Models, Inference & Training Python 545d 36543 - +652 +749
647 Wei-Shaw/sub2api ↗
Sub2API 一站式开源中转服务，让 Claude、Openai 、Gemini、Grok订阅统一接入，支持拼车共享，更高效分摊成本，原生工具无缝使用。
 LLM Models, Inference & Training Go 285d 43024 - +74 +748
81 ggml-org/llama.cpp ↗
LLM inference in C/C++
 LLM Models, Inference & Training C++ 1298d 129830 - +87 +741
1030 FareedKhan-dev/kimi-k3-in-c ↗
A 2.78-trillion-parameter Kimi K3 running inference on a single CPU in 8.24 GB of RAM. Portable C99: no BLAS, no framework, no GPU.
 LLM Models, Inference & Training C 58d 8781 - +44 +647
152 vllm-project/vllm ↗
A high-throughput and memory-efficient inference and serving engine for LLMs
 LLM Models, Inference & Training Python 1327d 92912 - +71 +555
926 lyogavin/airllm ↗
AirLLM 70B inference with single 4GB GPU
 LLM Models, Inference & Training Jupyter Notebook 1204d 35179 - +44 +532
1014 FlashML-org/FreeToken ↗
FreeToken brings datacenter-scale model serving to your desktop. Run massive models locally, fast and efficiently.
 LLM Models, Inference & Training Python 70d 13955 - +36 +505
43 ollama/ollama ↗
Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
 LLM Models, Inference & Training Go 1190d 181894 - +58 +501
376 BerriAI/litellm ↗
The fastest, litest AI Gateway. Rust core with Python SDK. Call 100+ LLM APIs in OpenAI (or native) format with cost tracking, guardrails, load balancing, and logging [Bedrock, Azure, OpenAI, Anthropic, OpenAI, VertexAI, vLLM, Nvidia NIM]
 LLM Models, Inference & Training Python 1160d 59825 - +67 +483
223 unslothai/unsloth ↗
Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4, MiniMax-H3, Gemma 4, FLUX and more.
 LLM Models, Inference & Training Python 1034d 76990 - +98 +450
520 QuantumNous/new-api ↗
A unified AI model hub for aggregation & distribution. It supports cross-converting various LLMs into OpenAI-compatible, Claude-compatible, or Gemini-compatible formats. A centralized gateway for personal and enterprise model management.
 LLM Models, Inference & Training Go 1053d 49040 - +42 +432
29 tensorflow/tensorflow ↗
An Open Source Machine Learning Framework for Everyone
 LLM Models, Inference & Training C++ 3979d 200603 - +26 +373
641 chatanywhere/GPT_API_free ↗
免费大模型API，支持免费调用GPT、DeepSeek等主流模型，免费额度10000点，每日刷新！另付费价格最低官方1-2折！
 LLM Models, Inference & Training 1254d 43342 - +51 +354
835 AlexsJones/llmfit ↗
Hundreds of models & providers. One command to find what runs on your hardware.
 LLM Models, Inference & Training Rust 225d 37287 - +63 +345
1087 yynxxxxx/Codex-X ↗
OpenAI Codex 桌面端/CLI 的可视化管理工具，具有Provider/API 切换、会话同步、提示词注入、Skills/MCP 管理、TOML 配置可视化的跨平台工具。
 LLM Models, Inference & Training Rust 87d 3988 - +21 +337
121 pytorch/pytorch ↗
Tensors and Dynamic neural networks in Python with strong GPU acceleration
 LLM Models, Inference & Training Python 3699d 103485 - +51 +330
50 huggingface/transformers ↗
🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.
LLM Models, Inference & Training Python 2891d 166790 - +42 +313
747 shiyu-coder/Kronos ↗
Kronos: A Foundation Model for the Language of Financial Markets
 LLM Models, Inference & Training Python 455d 39627 - +99 +308
1026 MiniMax-AI/MiniMax-H3 ↗
LLM Models, Inference & Training Python 60d 9350 - +41 +305
869 sgl-project/sglang ↗
SGLang is a high-performance serving framework for large language models and multimodal models.
 LLM Models, Inference & Training Python 995d 36558 - +52 +287
109 openai/whisper ↗
Robust Speech Recognition via Large-Scale Weak Supervision
 LLM Models, Inference & Training Python 1473d 109714 - +39 +273
769 google/langextract ↗
A Python library for extracting structured information from unstructured text using LLMs with precise source grounding and interactive visualization.
 LLM Models, Inference & Training Python 447d 38910 - +7 +271
804 OpenBMB/VoxCPM ↗
VoxCPM2: Tokenizer-Free TTS for Multilingual Speech Generation, Creative Voice Design, and True-to-Life Cloning
 LLM Models, Inference & Training Python 378d 38119 - +90 +265
989 SillyTavern/SillyTavern ↗
LLM Frontend for Power Users.
 LLM Models, Inference & Training JavaScript 1327d 33900 - +41 +257
793 stanfordnlp/dspy ↗
DSPy: The framework for programming—not prompting—language models
 LLM Models, Inference & Training Python 1358d 38410 - +21 +226
345 RVC-Boss/GPT-SoVITS ↗
1 min voice data can also be used to train a good TTS model! (few shot voice cloning)
 LLM Models, Inference & Training Python 988d 62239 - +16 +223
241 hiyouga/LlamaFactory ↗
Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)
 LLM Models, Inference & Training Python 1219d 75169 - +37 +211
783 RVC-Project/Retrieval-based-Voice-Conversion-WebUI ↗
Easily train a good VC model with voice data <= 10 mins!
 LLM Models, Inference & Training Python 1281d 38615 - +21 +191
441 ggml-org/whisper.cpp ↗
Port of OpenAI's Whisper model in C/C++
 LLM Models, Inference & Training C++ 1464d 53999 - +22 +157
328 karpathy/nanoGPT ↗
The simplest, fastest repository for training/finetuning medium-sized GPTs.
 LLM Models, Inference & Training Python 1371d 63437 - +17 +144
1269 syv-ai/HyperQwen ↗
Serve large Qwen models fast on the GPUs you actually own. Qwen3.8-27B on a single 24 GB card with vLLM: 127 tok/s single-user (381 when the answer quotes the prompt), ~1,035 tok/s at 64 concurrent, 150k-262k context. vLLM patches, requant pipeline, benchmarks.
 LLM Models, Inference & Training Python 44d 1755 - +16 +144
391 karpathy/nanochat ↗
The best ChatGPT that $100 can buy.
 LLM Models, Inference & Training Python 350d 58318 - +22 +115
611 janhq/jan ↗
Jan is an open source alternative to ChatGPT that runs 100% offline on your computer.
 LLM Models, Inference & Training Rust 1139d 44698 - +16 +113
815 myshell-ai/OpenVoice ↗
Instant voice cloning by MIT and MyShell. Audio foundation model.
 LLM Models, Inference & Training Python 1034d 37717 - +17 +111
735 HKUDS/LightRAG ↗
[EMNLP2025] LightRAG: Simple and Fast Retrieval-Augmented Generation
 LLM Models, Inference & Training Python 726d 39913 - +18 +110
514 mudler/LocalAI ↗
LocalAI is the open-source AI engine. Run any model - LLMs, vision, voice, image, video - on any hardware. No GPU required.
 LLM Models, Inference & Training Go 1290d 49315 - +11 +108
51 AUTOMATIC1111/stable-diffusion-webui ↗
Stable Diffusion web UI
 LLM Models, Inference & Training Python 1498d 165147 - 0 +102
481 roboflow/supervision ↗
We write your reusable computer vision tools. 💜
 LLM Models, Inference & Training Python 1400d 51071 - +4 +90
887 microsoft/graphrag ↗
A modular graph-based Retrieval-Augmented Generation (RAG) system
 LLM Models, Inference & Training Python 915d 36149 - +16 +89
453 lllyasviel/Fooocus ↗
Focus on prompting and generating
 LLM Models, Inference & Training Python 1146d 53224 - +19 +88
848 google-ai-edge/mediapipe ↗
Cross-platform, customizable ML solutions for live and streaming media.
 LLM Models, Inference & Training C++ 2664d 37114 - +14 +86
293 scikit-learn/scikit-learn ↗
scikit-learn: machine learning in Python
 LLM Models, Inference & Training Python 5886d 67416 - +10 +85
434 microsoft/VibeVoice ↗
Open-Source Frontier Voice AI
 LLM Models, Inference & Training Python 399d 54530 - +6 +81
548 exo-explore/exo ↗
Run frontier AI locally.
 LLM Models, Inference & Training Python 826d 47673 - +10 +79

LLM Models, Inference & Training, breakouts over 24 hours

#RepoCategory LanguageAge StarsNow24h7d
1165 YuJunZhiXue/dsh-purge ↗
DeepSeek Harness 破甲：让所有模型都能破甲，不同模型可换不同提示词；默认提示词面向国模「小码酱」。Jailbreak for every model — swap prompts per model. 求 Star 收藏 ⭐
 LLM Models, Inference & Training JavaScript 40d 2591 - +95 +825
1240 yibie/awesome-jev ↗
A curated list of public projects, integrations, and discussions built on Jev — TypeSafe AI's System One model for typed decisions.
 LLM Models, Inference & Training Python 11d 1941 - +69 -
1002 NandhaKishorM/laya ↗
Non-autoregressive System 1 decision engine. Typed choice, score and yes/no decisions over any text in a single forward pass, in 100+ languages, with a router that picks the right checkpoint per request.
 LLM Models, Inference & Training Python 11d 27941 - +875 +18337
1035 jaredpalmer/kev ↗
Jev-like family of decision models built on top of Qwen3.5/3.8 you can train and run on your own
 LLM Models, Inference & Training Python 11d 7713 - +208 +5416
1239 Taichu-AI/ZDTaichu5.0-9B ↗
LLM Models, Inference & Training Python 25d 1942 - +45 -
870 VectifyAI/PageIndex ↗
📑 PageIndex: Document Index for Vectorless, Reasoning-based RAG
 LLM Models, Inference & Training Python 545d 36543 - +652 +749
1291 QwenLM/Qwen-Image-2.1 ↗
Qwen's most powerful open-source image generation model
 LLM Models, Inference & Training Python 15d 1609 - +27 -
1048 mizorewww/laya-mlx ↗
Native MLX runtime for Laya typed decision models — 7–14 ms short decisions on M3 Max. No text generation, PyTorch, or cloud API.
 LLM Models, Inference & Training Python 9d 6593 - +78 +3018
1269 syv-ai/HyperQwen ↗
Serve large Qwen models fast on the GPUs you actually own. Qwen3.8-27B on a single 24 GB card with vLLM: 127 tok/s single-user (381 when the answer quotes the prompt), ~1,035 tok/s at 64 concurrent, 150k-262k context. vLLM patches, requant pipeline, benchmarks.
 LLM Models, Inference & Training Python 44d 1755 - +16 +144
1087 yynxxxxx/Codex-X ↗
OpenAI Codex 桌面端/CLI 的可视化管理工具，具有Provider/API 切换、会话同步、提示词注入、Skills/MCP 管理、TOML 配置可视化的跨平台工具。
 LLM Models, Inference & Training Rust 87d 3988 - +21 +337
1030 FareedKhan-dev/kimi-k3-in-c ↗
A 2.78-trillion-parameter Kimi K3 running inference on a single CPU in 8.24 GB of RAM. Portable C99: no BLAS, no framework, no GPU.
 LLM Models, Inference & Training C 58d 8781 - +44 +647
798 JustVugg/colibri ↗
Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦
 LLM Models, Inference & Training C 89d 38193 - +170 +1373
1026 MiniMax-AI/MiniMax-H3 ↗
LLM Models, Inference & Training Python 60d 9350 - +41 +305
1253 jamesob/local-llm ↗
Everything I know about running LLMs locally
 LLM Models, Inference & Training Shell 87d 1853 - +5 +12
1014 FlashML-org/FreeToken ↗
FreeToken brings datacenter-scale model serving to your desktop. Run massive models locally, fast and efficiently.
 LLM Models, Inference & Training Python 70d 13955 - +36 +505
747 shiyu-coder/Kronos ↗
Kronos: A Foundation Model for the Language of Financial Markets
 LLM Models, Inference & Training Python 455d 39627 - +99 +308
1278 Tencent/WeMM-Embedding ↗
WeMM-Embedding is a family of universal multimodal embedding models by the WeChat Vision Team at Tencent, supporting multimodal understanding and retrieval.
 LLM Models, Inference & Training Python 35d 1689 - +4 +45
804 OpenBMB/VoxCPM ↗
VoxCPM2: Tokenizer-Free TTS for Multilingual Speech Generation, Creative Voice Design, and True-to-Life Cloning
 LLM Models, Inference & Training Python 378d 38119 - +90 +265
986 google-research/timesfm ↗
TimesFM (Time Series Foundation Model) is a pretrained time-series foundation model developed by Google Research for time-series forecasting.
 LLM Models, Inference & Training Python 882d 33953 - +61 -
647 Wei-Shaw/sub2api ↗
Sub2API 一站式开源中转服务，让 Claude、Openai 、Gemini、Grok订阅统一接入，支持拼车共享，更高效分摊成本，原生工具无缝使用。
 LLM Models, Inference & Training Go 285d 43024 - +74 +748
835 AlexsJones/llmfit ↗
Hundreds of models & providers. One command to find what runs on your hardware.
 LLM Models, Inference & Training Rust 225d 37287 - +63 +345
869 sgl-project/sglang ↗
SGLang is a high-performance serving framework for large language models and multimodal models.
 LLM Models, Inference & Training Python 995d 36558 - +52 +287
223 unslothai/unsloth ↗
Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4, MiniMax-H3, Gemma 4, FLUX and more.
 LLM Models, Inference & Training Python 1034d 76990 - +98 +450
926 lyogavin/airllm ↗
AirLLM 70B inference with single 4GB GPU
 LLM Models, Inference & Training Jupyter Notebook 1204d 35179 - +44 +532
1172 sqliteai/warp ↗
Run the full 2.78-trillion-parameter Kimi K3 model, DeepSeek V4.1 Flash or GLM-5.3-Flash beyond available RAM by streaming activated weights directly from NVMe. A dependency-free, embeddable C inference engine.
 LLM Models, Inference & Training C 62d 2457 - +3 +21
989 SillyTavern/SillyTavern ↗
LLM Frontend for Power Users.
 LLM Models, Inference & Training JavaScript 1327d 33900 - +41 +257
641 chatanywhere/GPT_API_free ↗
免费大模型API，支持免费调用GPT、DeepSeek等主流模型，免费额度10000点，每日刷新！另付费价格最低官方1-2折！
 LLM Models, Inference & Training 1254d 43342 - +51 +354
376 BerriAI/litellm ↗
The fastest, litest AI Gateway. Rust core with Python SDK. Call 100+ LLM APIs in OpenAI (or native) format with cost tracking, guardrails, load balancing, and logging [Bedrock, Azure, OpenAI, Anthropic, OpenAI, VertexAI, vLLM, Nvidia NIM]
 LLM Models, Inference & Training Python 1160d 59825 - +67 +483
339 jingyaogong/minimind ↗
🧠 Train a 64M-parameter LLM from scratch in just 2h!
 LLM Models, Inference & Training Python 794d 62862 - +67 +840
1045 drumih/turbo-fieldfare ↗
Gemma 4 26B-A4B inference in ~2 GB of RAM on any M-series MacBook
 LLM Models, Inference & Training Swift 73d 6841 - +6 +46
520 QuantumNous/new-api ↗
A unified AI model hub for aggregation & distribution. It supports cross-converting various LLMs into OpenAI-compatible, Claude-compatible, or Gemini-compatible formats. A centralized gateway for personal and enterprise model management.
 LLM Models, Inference & Training Go 1053d 49040 - +42 +432
152 vllm-project/vllm ↗
A high-throughput and memory-efficient inference and serving engine for LLMs
 LLM Models, Inference & Training Python 1327d 92912 - +71 +555
81 ggml-org/llama.cpp ↗
LLM inference in C/C++
 LLM Models, Inference & Training C++ 1298d 129830 - +87 +741
1029 MoonshotAI/Kimi-K3 ↗
Open Frontier Intelligence
 LLM Models, Inference & Training 63d 8876 - +5 +40
793 stanfordnlp/dspy ↗
DSPy: The framework for programming—not prompting—language models
 LLM Models, Inference & Training Python 1358d 38410 - +21 +226
783 RVC-Project/Retrieval-based-Voice-Conversion-WebUI ↗
Easily train a good VC model with voice data <= 10 mins!
 LLM Models, Inference & Training Python 1281d 38615 - +21 +191
121 pytorch/pytorch ↗
Tensors and Dynamic neural networks in Python with strong GPU acceleration
 LLM Models, Inference & Training Python 3699d 103485 - +51 +330
241 hiyouga/LlamaFactory ↗
Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)
 LLM Models, Inference & Training Python 1219d 75169 - +37 +211
735 HKUDS/LightRAG ↗
[EMNLP2025] LightRAG: Simple and Fast Retrieval-Augmented Generation
 LLM Models, Inference & Training Python 726d 39913 - +18 +110
815 myshell-ai/OpenVoice ↗
Instant voice cloning by MIT and MyShell. Audio foundation model.
 LLM Models, Inference & Training Python 1034d 37717 - +17 +111
887 microsoft/graphrag ↗
A modular graph-based Retrieval-Augmented Generation (RAG) system
 LLM Models, Inference & Training Python 915d 36149 - +16 +89
441 ggml-org/whisper.cpp ↗
Port of OpenAI's Whisper model in C/C++
 LLM Models, Inference & Training C++ 1464d 53999 - +22 +157
391 karpathy/nanochat ↗
The best ChatGPT that $100 can buy.
 LLM Models, Inference & Training Python 350d 58318 - +22 +115
848 google-ai-edge/mediapipe ↗
Cross-platform, customizable ML solutions for live and streaming media.
 LLM Models, Inference & Training C++ 2664d 37114 - +14 +86
611 janhq/jan ↗
Jan is an open source alternative to ChatGPT that runs 100% offline on your computer.
 LLM Models, Inference & Training Rust 1139d 44698 - +16 +113
453 lllyasviel/Fooocus ↗
Focus on prompting and generating
 LLM Models, Inference & Training Python 1146d 53224 - +19 +88
109 openai/whisper ↗
Robust Speech Recognition via Large-Scale Weak Supervision
 LLM Models, Inference & Training Python 1473d 109714 - +39 +273
859 xinntao/Real-ESRGAN ↗
Real-ESRGAN aims at developing Practical Algorithms for General Image/Video Restoration.
 LLM Models, Inference & Training Python 1898d 36939 - +13 +72
43 ollama/ollama ↗
Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
 LLM Models, Inference & Training Go 1190d 181894 - +58 +501
328 karpathy/nanoGPT ↗
The simplest, fastest repository for training/finetuning medium-sized GPTs.
 LLM Models, Inference & Training Python 1371d 63437 - +17 +144

LLM Models, Inference & Training, breakouts over 7 days

#RepoCategory LanguageAge StarsNow24h7d
1035 jaredpalmer/kev ↗
Jev-like family of decision models built on top of Qwen3.5/3.8 you can train and run on your own
 LLM Models, Inference & Training Python 11d 7713 - +208 +5416
1002 NandhaKishorM/laya ↗
Non-autoregressive System 1 decision engine. Typed choice, score and yes/no decisions over any text in a single forward pass, in 100+ languages, with a router that picks the right checkpoint per request.
 LLM Models, Inference & Training Python 11d 27941 - +875 +18337
1048 mizorewww/laya-mlx ↗
Native MLX runtime for Laya typed decision models — 7–14 ms short decisions on M3 Max. No text generation, PyTorch, or cloud API.
 LLM Models, Inference & Training Python 9d 6593 - +78 +3018
1165 YuJunZhiXue/dsh-purge ↗
DeepSeek Harness 破甲：让所有模型都能破甲，不同模型可换不同提示词；默认提示词面向国模「小码酱」。Jailbreak for every model — swap prompts per model. 求 Star 收藏 ⭐
 LLM Models, Inference & Training JavaScript 40d 2591 - +95 +825
1087 yynxxxxx/Codex-X ↗
OpenAI Codex 桌面端/CLI 的可视化管理工具，具有Provider/API 切换、会话同步、提示词注入、Skills/MCP 管理、TOML 配置可视化的跨平台工具。
 LLM Models, Inference & Training Rust 87d 3988 - +21 +337
1269 syv-ai/HyperQwen ↗
Serve large Qwen models fast on the GPUs you actually own. Qwen3.8-27B on a single 24 GB card with vLLM: 127 tok/s single-user (381 when the answer quotes the prompt), ~1,035 tok/s at 64 concurrent, 150k-262k context. vLLM patches, requant pipeline, benchmarks.
 LLM Models, Inference & Training Python 44d 1755 - +16 +144
1030 FareedKhan-dev/kimi-k3-in-c ↗
A 2.78-trillion-parameter Kimi K3 running inference on a single CPU in 8.24 GB of RAM. Portable C99: no BLAS, no framework, no GPU.
 LLM Models, Inference & Training C 58d 8781 - +44 +647
1014 FlashML-org/FreeToken ↗
FreeToken brings datacenter-scale model serving to your desktop. Run massive models locally, fast and efficiently.
 LLM Models, Inference & Training Python 70d 13955 - +36 +505
798 JustVugg/colibri ↗
Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦
 LLM Models, Inference & Training C 89d 38193 - +170 +1373
1026 MiniMax-AI/MiniMax-H3 ↗
LLM Models, Inference & Training Python 60d 9350 - +41 +305
1278 Tencent/WeMM-Embedding ↗
WeMM-Embedding is a family of universal multimodal embedding models by the WeChat Vision Team at Tencent, supporting multimodal understanding and retrieval.
 LLM Models, Inference & Training Python 35d 1689 - +4 +45
870 VectifyAI/PageIndex ↗
📑 PageIndex: Document Index for Vectorless, Reasoning-based RAG
 LLM Models, Inference & Training Python 545d 36543 - +652 +749
647 Wei-Shaw/sub2api ↗
Sub2API 一站式开源中转服务，让 Claude、Openai 、Gemini、Grok订阅统一接入，支持拼车共享，更高效分摊成本，原生工具无缝使用。
 LLM Models, Inference & Training Go 285d 43024 - +74 +748
926 lyogavin/airllm ↗
AirLLM 70B inference with single 4GB GPU
 LLM Models, Inference & Training Jupyter Notebook 1204d 35179 - +44 +532
339 jingyaogong/minimind ↗
🧠 Train a 64M-parameter LLM from scratch in just 2h!
 LLM Models, Inference & Training Python 794d 62862 - +67 +840
835 AlexsJones/llmfit ↗
Hundreds of models & providers. One command to find what runs on your hardware.
 LLM Models, Inference & Training Rust 225d 37287 - +63 +345
520 QuantumNous/new-api ↗
A unified AI model hub for aggregation & distribution. It supports cross-converting various LLMs into OpenAI-compatible, Claude-compatible, or Gemini-compatible formats. A centralized gateway for personal and enterprise model management.
 LLM Models, Inference & Training Go 1053d 49040 - +42 +432
1172 sqliteai/warp ↗
Run the full 2.78-trillion-parameter Kimi K3 model, DeepSeek V4.1 Flash or GLM-5.3-Flash beyond available RAM by streaming activated weights directly from NVMe. A dependency-free, embeddable C inference engine.
 LLM Models, Inference & Training C 62d 2457 - +3 +21
641 chatanywhere/GPT_API_free ↗
免费大模型API，支持免费调用GPT、DeepSeek等主流模型，免费额度10000点，每日刷新！另付费价格最低官方1-2折！
 LLM Models, Inference & Training 1254d 43342 - +51 +354
376 BerriAI/litellm ↗
The fastest, litest AI Gateway. Rust core with Python SDK. Call 100+ LLM APIs in OpenAI (or native) format with cost tracking, guardrails, load balancing, and logging [Bedrock, Azure, OpenAI, Anthropic, OpenAI, VertexAI, vLLM, Nvidia NIM]
 LLM Models, Inference & Training Python 1160d 59825 - +67 +483
869 sgl-project/sglang ↗
SGLang is a high-performance serving framework for large language models and multimodal models.
 LLM Models, Inference & Training Python 995d 36558 - +52 +287
747 shiyu-coder/Kronos ↗
Kronos: A Foundation Model for the Language of Financial Markets
 LLM Models, Inference & Training Python 455d 39627 - +99 +308
989 SillyTavern/SillyTavern ↗
LLM Frontend for Power Users.
 LLM Models, Inference & Training JavaScript 1327d 33900 - +41 +257
769 google/langextract ↗
A Python library for extracting structured information from unstructured text using LLMs with precise source grounding and interactive visualization.
 LLM Models, Inference & Training Python 447d 38910 - +7 +271
804 OpenBMB/VoxCPM ↗
VoxCPM2: Tokenizer-Free TTS for Multilingual Speech Generation, Creative Voice Design, and True-to-Life Cloning
 LLM Models, Inference & Training Python 378d 38119 - +90 +265
1045 drumih/turbo-fieldfare ↗
Gemma 4 26B-A4B inference in ~2 GB of RAM on any M-series MacBook
 LLM Models, Inference & Training Swift 73d 6841 - +6 +46
1253 jamesob/local-llm ↗
Everything I know about running LLMs locally
 LLM Models, Inference & Training Shell 87d 1853 - +5 +12
152 vllm-project/vllm ↗
A high-throughput and memory-efficient inference and serving engine for LLMs
 LLM Models, Inference & Training Python 1327d 92912 - +71 +555
793 stanfordnlp/dspy ↗
DSPy: The framework for programming—not prompting—language models
 LLM Models, Inference & Training Python 1358d 38410 - +21 +226
223 unslothai/unsloth ↗
Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4, MiniMax-H3, Gemma 4, FLUX and more.
 LLM Models, Inference & Training Python 1034d 76990 - +98 +450
81 ggml-org/llama.cpp ↗
LLM inference in C/C++
 LLM Models, Inference & Training C++ 1298d 129830 - +87 +741
783 RVC-Project/Retrieval-based-Voice-Conversion-WebUI ↗
Easily train a good VC model with voice data <= 10 mins!
 LLM Models, Inference & Training Python 1281d 38615 - +21 +191
1029 MoonshotAI/Kimi-K3 ↗
Open Frontier Intelligence
 LLM Models, Inference & Training 63d 8876 - +5 +40
345 RVC-Boss/GPT-SoVITS ↗
1 min voice data can also be used to train a good TTS model! (few shot voice cloning)
 LLM Models, Inference & Training Python 988d 62239 - +16 +223
121 pytorch/pytorch ↗
Tensors and Dynamic neural networks in Python with strong GPU acceleration
 LLM Models, Inference & Training Python 3699d 103485 - +51 +330
815 myshell-ai/OpenVoice ↗
Instant voice cloning by MIT and MyShell. Audio foundation model.
 LLM Models, Inference & Training Python 1034d 37717 - +17 +111
441 ggml-org/whisper.cpp ↗
Port of OpenAI's Whisper model in C/C++
 LLM Models, Inference & Training C++ 1464d 53999 - +22 +157
241 hiyouga/LlamaFactory ↗
Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)
 LLM Models, Inference & Training Python 1219d 75169 - +37 +211
735 HKUDS/LightRAG ↗
[EMNLP2025] LightRAG: Simple and Fast Retrieval-Augmented Generation
 LLM Models, Inference & Training Python 726d 39913 - +18 +110
43 ollama/ollama ↗
Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
 LLM Models, Inference & Training Go 1190d 181894 - +58 +501
611 janhq/jan ↗
Jan is an open source alternative to ChatGPT that runs 100% offline on your computer.
 LLM Models, Inference & Training Rust 1139d 44698 - +16 +113
109 openai/whisper ↗
Robust Speech Recognition via Large-Scale Weak Supervision
 LLM Models, Inference & Training Python 1473d 109714 - +39 +273
887 microsoft/graphrag ↗
A modular graph-based Retrieval-Augmented Generation (RAG) system
 LLM Models, Inference & Training Python 915d 36149 - +16 +89
848 google-ai-edge/mediapipe ↗
Cross-platform, customizable ML solutions for live and streaming media.
 LLM Models, Inference & Training C++ 2664d 37114 - +14 +86
328 karpathy/nanoGPT ↗
The simplest, fastest repository for training/finetuning medium-sized GPTs.
 LLM Models, Inference & Training Python 1371d 63437 - +17 +144
514 mudler/LocalAI ↗
LocalAI is the open-source AI engine. Run any model - LLMs, vision, voice, image, video - on any hardware. No GPU required.
 LLM Models, Inference & Training Go 1290d 49315 - +11 +108
391 karpathy/nanochat ↗
The best ChatGPT that $100 can buy.
 LLM Models, Inference & Training Python 350d 58318 - +22 +115
859 xinntao/Real-ESRGAN ↗
Real-ESRGAN aims at developing Practical Algorithms for General Image/Video Restoration.
 LLM Models, Inference & Training Python 1898d 36939 - +13 +72
949 huggingface/diffusers ↗
🤗 Diffusers: State-of-the-art diffusion models for image, video, and audio generation in PyTorch.
 LLM Models, Inference & Training Python 1582d 34628 - +6 +65
50 huggingface/transformers ↗
🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.
LLM Models, Inference & Training Python 2891d 166790 - +42 +313

LLM Models, Inference & Training, new and rising

#RepoCategory LanguageAge StarsNow24h7d
798 JustVugg/colibri ↗
Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦
 LLM Models, Inference & Training C 89d 38193 - +170 +1373
1002 NandhaKishorM/laya ↗
Non-autoregressive System 1 decision engine. Typed choice, score and yes/no decisions over any text in a single forward pass, in 100+ languages, with a router that picks the right checkpoint per request.
 LLM Models, Inference & Training Python 11d 27941 - +875 +18337
1014 FlashML-org/FreeToken ↗
FreeToken brings datacenter-scale model serving to your desktop. Run massive models locally, fast and efficiently.
 LLM Models, Inference & Training Python 70d 13955 - +36 +505
1018 pathwaycom/arc-task-gen ↗
Generates original ARC-AGI-1-style tasks distribution-matched to the public eval set.
 LLM Models, Inference & Training Python 55d 11113 - -6 -57
1026 MiniMax-AI/MiniMax-H3 ↗
LLM Models, Inference & Training Python 60d 9350 - +41 +305
1029 MoonshotAI/Kimi-K3 ↗
Open Frontier Intelligence
 LLM Models, Inference & Training 63d 8876 - +5 +40
1030 FareedKhan-dev/kimi-k3-in-c ↗
A 2.78-trillion-parameter Kimi K3 running inference on a single CPU in 8.24 GB of RAM. Portable C99: no BLAS, no framework, no GPU.
 LLM Models, Inference & Training C 58d 8781 - +44 +647
1035 jaredpalmer/kev ↗
Jev-like family of decision models built on top of Qwen3.5/3.8 you can train and run on your own
 LLM Models, Inference & Training Python 11d 7713 - +208 +5416
1045 drumih/turbo-fieldfare ↗
Gemma 4 26B-A4B inference in ~2 GB of RAM on any M-series MacBook
 LLM Models, Inference & Training Swift 73d 6841 - +6 +46
1048 mizorewww/laya-mlx ↗
Native MLX runtime for Laya typed decision models — 7–14 ms short decisions on M3 Max. No text generation, PyTorch, or cloud API.
 LLM Models, Inference & Training Python 9d 6593 - +78 +3018
1087 yynxxxxx/Codex-X ↗
OpenAI Codex 桌面端/CLI 的可视化管理工具，具有Provider/API 切换、会话同步、提示词注入、Skills/MCP 管理、TOML 配置可视化的跨平台工具。
 LLM Models, Inference & Training Rust 87d 3988 - +21 +337
1165 YuJunZhiXue/dsh-purge ↗
DeepSeek Harness 破甲：让所有模型都能破甲，不同模型可换不同提示词；默认提示词面向国模「小码酱」。Jailbreak for every model — swap prompts per model. 求 Star 收藏 ⭐
 LLM Models, Inference & Training JavaScript 40d 2591 - +95 +825
1172 sqliteai/warp ↗
Run the full 2.78-trillion-parameter Kimi K3 model, DeepSeek V4.1 Flash or GLM-5.3-Flash beyond available RAM by streaming activated weights directly from NVMe. A dependency-free, embeddable C inference engine.
 LLM Models, Inference & Training C 62d 2457 - +3 +21
1239 Taichu-AI/ZDTaichu5.0-9B ↗
LLM Models, Inference & Training Python 25d 1942 - +45 -
1240 yibie/awesome-jev ↗
A curated list of public projects, integrations, and discussions built on Jev — TypeSafe AI's System One model for typed decisions.
 LLM Models, Inference & Training Python 11d 1941 - +69 -
1253 jamesob/local-llm ↗
Everything I know about running LLMs locally
 LLM Models, Inference & Training Shell 87d 1853 - +5 +12
1269 syv-ai/HyperQwen ↗
Serve large Qwen models fast on the GPUs you actually own. Qwen3.8-27B on a single 24 GB card with vLLM: 127 tok/s single-user (381 when the answer quotes the prompt), ~1,035 tok/s at 64 concurrent, 150k-262k context. vLLM patches, requant pipeline, benchmarks.
 LLM Models, Inference & Training Python 44d 1755 - +16 +144
1278 Tencent/WeMM-Embedding ↗
WeMM-Embedding is a family of universal multimodal embedding models by the WeChat Vision Team at Tencent, supporting multimodal understanding and retrieval.
 LLM Models, Inference & Training Python 35d 1689 - +4 +45
1291 QwenLM/Qwen-Image-2.1 ↗
Qwen's most powerful open-source image generation model
 LLM Models, Inference & Training Python 15d 1609 - +27 -