NodeShift Blog

Featured blog post

September 10, 2025

How to Install & Run ERNIE-4.5-21B-A3B-Thinking Locally?

A 21B-parameter text MoE (Mixture-of-Experts) model with 3B activated params/token, post-trained for deep reasoning. It adds stronger tool use, long-context (131,072 tokens), and higher pass@1/accuracy on math/logic, coding, science, and academic benchmarks. Weights are released in Transformer-style (PyTorch) with BF16 / FP32, and it can be run via FastDeploy (recommended) or standard transformers. Function-calling is supported; vLLM parsers for reasoning/tool calls are in progress. Key config: 28 layers, 20 Q heads / 4 KV heads, 64 text experts (6 active), 2 shared experts. License: Apache-2.0.

All blog posts

September 10, 2025

How to Install & Run K2-Think Locally?

K2-Think is a 32B open-weights reasoning model focused on tough math/logic, code, and science tasks. It’s trained for long chain-of-thought and integrates reinforcement learning with verifiable rewards and agentic planning. Despite its size, it targets high efficiency: the team reports ~2,000 tok/s on Cerebras WSE with speculative decoding (vs. ~200 tok/s on typical H100/H200 setups), and strong scores on AIME’24/’25, HMMT’25, OMNI-Math-HARD, GPQA-Diamond, and LiveCodeBench. Weights are Apache-2.0 and served on Hugging Face.

September 9, 2025

How to Install & Run Apertus: The Massive Multilingual AI Model Supporting 1,800+ Languages

The AI landscape has been dominated by a handful of large language models, many of which operate as “black boxes” with hidden data and opaque training methods. But Apertus enters the AI space as the state-of-the-art model that is completely transparent, from its training data to its core architecture. Apertus, that comes in both 8B and 70B parameter variants, distinguishes itself not just with its size, but with its commitment to radical transparency and massive multilingualism. It was pre-trained on an unprecedented 15 trillion tokens, with over 40% of the data in languages other than English, providing native support for over 1,800 languages, a milestone that makes it uniquely valuable for global applications and under-resourced linguistic communities. Unlike many models that only offer their weights, Apertus provides all the scientific artifacts from its development cycle, including data preparation scripts, training code, and evaluation suites, allowing for transparent audits and community-driven extension. This model is a foundational blueprint for the future of ethical, compliant, and inclusive AI.

September 9, 2025

How to Install & Run R-4B: Auto-Thinking Model Locally?

R-4B is a multimodal large language model designed to introduce general-purpose auto-thinking. Unlike traditional models that either always perform step-by-step reasoning or skip it entirely, R-4B can adaptively switch between thinking and non-thinking modes depending on task complexity. This is achieved through its Bi-mode Annealing training (to build both capabilities) and Bi-mode Policy Optimization (to dynamically balance them during inference). This flexibility allows R-4B to handle everything from quick Q&A to complex logical or scientific reasoning while keeping efficiency high. With recent integration into vLLM, R-4B also enables fast, scalable deployments and exposes a simple API for manual or automatic control over its “thinking mode.” It already tops multiple OpenCompass multimodal leaderboards, making it one of the most advanced open-source reasoning-capable MLLMs under 20B parameters.

September 8, 2025

How to Install HunyuanWorld-Voyager: Create Stunning 3D Images & Videos from a Single Image

Have you ever wanted to create and explore vast, consistent 3D worlds from a single image? While previous models like HunyuanWorld 1.0 have made strides in explorable 3D world generation, they often struggle with occluded views and limited exploration ranges. This is where HunyuanWorld-Voyager comes in, a groundbreaking video diffusion framework that shatters these limitations by generating world-consistent 3D point-cloud sequences. Voyager addresses these challenges head-on, offering a world-consistent video diffusion framework capable of generating coherent 3D point-cloud sequences from a single image along a user-defined camera path. By integrating world-consistent video diffusion, long-range world exploration, and a scalable data engine, Voyager ensures end-to-end scene generation and reconstruction without relying on traditional 3D reconstruction pipelines. Its innovative architecture jointly generates aligned RGB and depth sequences, leverages an efficient world cache with point culling for iterative exploration, and automatically predicts camera poses and metric depths, delivering unmatched visual fidelity and geometric accuracy.

September 8, 2025

How to Install & Run Microsoft Kosmos-2.5 Locally?

Kosmos-2.5 is Microsoft’s multimodal “literate” model for reading text-heavy images (receipts, invoices, forms, docs). It does two things out of the box using task prompts: (a) OCR with spatially-aware text blocks (text + bounding boxes) via , and (b) image→Markdown conversion via . It’s implemented in Transformers (supported from v4.56+) with ready-to-run Python snippets, and the paper details the shared decoder-only architecture and doc-understanding focus

September 6, 2025

How to Install & Run EmbeddingGemma-300m Locally?

EmbeddingGemma-300M is Google DeepMind’s lightweight, multilingual (100+ languages) embedding model built on Gemma 3/T5Gemma foundations. It outputs 768-dim vectors (with Matryoshka down-projections to 512/256/128) optimized for retrieval, classification, clustering, semantic similarity, QA, and code retrieval. It’s designed for low-resource / on-device use, loads via SentenceTransformers, and does not support float16—use FP32 or bfloat16.

September 4, 2025

Build Advanced Speech-to-Speech Systems with Step-Audio 2

Step-Audio 2 is an advanced, end-to-end multi-modal large language model designed to transform how we interact with audio. It goes beyond simple transcription, offering a deep, nuanced understanding of speech and audio environments. Think of a model that not only comprehends the words in a conversation but also grasps the para-linguistic things like tone and emotion, and even recognizes non-vocal information like background noise. This advanced capability allows for highly sophisticated audio analysis, making it a promising solution for complex tasks like customer service automation, content analysis, and real-time conversational agents. Step-Audio 2’s unique ability to reason about and act upon this rich acoustic data places it at the forefront of audio AI, promising a level of intelligence and contextual awareness that has been a challenge for previous models.

September 4, 2025

How to Install & Run ByteDance USO Locally?

USO (Unified Style–Subject Optimized) from ByteDance unifies style-driven and subject-driven image generation in one framework. It’s trained on triplets (content image, style image, stylized image) and uses a disentangled learning scheme—style-alignment + content–style disentanglement—plus Style Reward Learning (SRL) to boost fidelity. The team also releases USO-Bench, the first benchmark that jointly scores style similarity and subject fidelity; USO reports SOTA among open-source models on both axes. Inference runs on top of FLUX.1 (with AE, T5, CLIP) and lightweight USO adapters (LoRA + projector).

September 3, 2025

How to Install & Run MiniCPM-V-4_5 Locally?

MiniCPM-V 4.5 is the latest milestone in the MiniCPM Vision-Language series by OpenBMB. Built on Qwen3-8B with a SigLIP2-400M vision encoder, this model delivers GPT-4o-level multimodal performance with only ~8.7B parameters. It outperforms models like GPT-4o-latest and Gemini 2.0 Pro in OCR, document parsing, and video understanding—all while being lightweight enough to run on your phone. Key highlights: Best-in-class MLLM under 30B: Tops OpenCompass with a 77.0 avg score 96× video token compression via 3D-Resampler for long & high-FPS video Switchable fast vs. deep thinking modes via hybrid RL training Top-tier OCR & document reasoning using UHD vision input Multilingual understanding (30+ languages) and trustworthy outputs (via RLAIF-V) It supports various deployment methods: Llama.cpp, Ollama, vLLM, SGLang, Gradio UI, and even iOS demos—making it one of the most accessible MLLMs ever.

September 2, 2025

How to Install Hunyuan-MT-7B Locally: Groundbreaking Machine Translation Model for 33 Languages

If you’re also struggling with broken hallucinating translation tools or looking for more powerful model running right on your own machine, you’re going to love this. Tencent has launched Hunyuan-MT-7B, a translation model that’s been making waves in the AI community. It’s designed to handle translations across 33 languages, even covering several of regional ethnic and minority languages like Marathi, Bengali, Cantonese, Polish and many more spoken in several countries like India, China, Europe and worldwide, which is pretty rare in open-source models. And it doesn’t just stop at being versatile, the performance is off the charts. At the WMT25 competition, this model grabbed first place in 30 out of 31 language categories it entered. That’s the kind of track record you usually only see in massive, closed-source systems, but now it’s available for anyone to try. Plus, it comes with a sibling, Hunyuan-MT-Chimera-7B, which is the industry’s first open-source ensemble translation model. Think of it as a super-refiner that takes multiple translation outputs and blends them into the best possible version. Together, these models bring speed, accuracy, and quality that’s hard to beat.

Trusted by Thousands of Cloud Professionals

Ready to build
with us?

The ideal way for organizations young and old to ease their way into the distributed and affordable cloud at their own pace.

Stay Tuned!

Stay up to date with the latest updates, news, and hotfixes for our product.

NodeShift creates a vital link between developers and affordable cloud.

Switch theme

English (EN)
Arabic (AR)
Chinese (ZH-CN)
German (DE)
Korean (KO)
Russian (RU)
French (FR)
Spanish (ES)
Portuguese (PT)
Japanese (JA)

JavaScript is disabled in your browser. For a better experience, please enable JavaScript.Learn how to enable JavaScript.