All LLM Models

Browse 1475 LLM models with VRAM requirements, quantization options, and hardware compatibility.

Featured only

Understanding LLM VRAM Requirements

How much VRAM you need depends on the model size and quantization level. Quantization reduces the precision of model weights, trading small quality losses for significantly lower VRAM usage. For example, a 7B parameter model needs ~14 GB at FP16 but only ~4 GB at Q4_K_M quantization.

GLM 5.2

Z.ai · 753.3B · runs from 211.4 GB

GLM 5.2, released by Z.ai in June 2026, is a mixture-of-experts language model in the GLM 5 family with roughly 753 billion total parameters. As an MoE design, it activates only a subset of those parameters for any given token, which helps keep inference more efficient than a dense model of the same total size, even though it remains a very large checkpoint overall. The model offers a 1M token context window, suited to long-document and long-context workloads, and is released under the MIT license, one of the more permissive terms available for local and commercial use. Given its total parameter count, running GLM 5.2 locally needs multi-GPU or server-class hardware; most users will access it through a hosted endpoint rather than on a single machine.

Muse Glimmer 30B

Meta · 29.8B · runs from 8.7 GB

Muse Glimmer 30B is Meta's 30-billion-parameter dense vision-language model, purpose-built for local, always-on AI agents rather than general chat. It pairs a text decoder with a dedicated image encoder for reasoning over screenshots, charts, and documents, and includes native tool-calling with a separate reasoning channel so it can plan multi-step actions and recover from failures. At this size, local inference calls for quantization and a capable GPU; it fits a single high-end 24-32GB-class card, matching Meta's goal of running it entirely on consumer machines. The model supports a 131K token context window, enough for extended agent sessions. It is released under the Apache 2.0 license, allowing unrestricted commercial and research use. Published in August 2026, it is Meta's first open-weight model since Llama 4, and ships without the Llama licenses' monthly-active-user cap.

Llama 3.2 3B Instruct

Meta · 3.2B · runs from 1.0 GB

Meta Llama 3.2 3B Instruct is a 3-billion parameter instruction-tuned model from Meta's Llama 3.2 release, designed for efficient local inference on resource-constrained hardware. It supports a 128K token context window and is optimized for conversational AI, summarization, and general assistant tasks. Despite its small footprint, Llama 3.2 3B Instruct delivers competitive performance for its size class and can run on GPUs with as little as 4GB of VRAM when quantized. It is released under the Llama 3.2 Community License and is a practical choice for edge deployment and lightweight local inference.

Qwen3 30B A3B Instruct 2507

Alibaba · 30.5B · runs from 13.4 GB

Qwen3 30B A3B Instruct 2507 is a July 2025 updated mixture-of-experts model from Alibaba with 30 billion total parameters but only around 3 billion active during inference. This MoE architecture gives it a remarkably small memory and compute footprint relative to its total parameter count, letting users run a model with broad knowledge on mid-range hardware. The 2507 instruct refresh improves alignment and instruction-following quality over the original release. Because only a fraction of the weights are active at any given time, this model can often run on a single consumer GPU with 8 GB or more of VRAM when quantized, making it an excellent choice for users who want strong chat performance without heavyweight hardware.

DeepSeek R1 0528 Qwen3 8B

DeepSeek · 8.2B · runs from 2.9 GB

With about 8 billion parameters, DeepSeek R1 0528 Qwen3 8B is a dated May 2025 release in DeepSeek's R1 line, built on a Qwen3-based architecture rather than DeepSeek's own base model. It supports explicit step-by-step reasoning, generating intermediate reasoning traces before producing a final answer, which sets it apart from standard instruction-tuned chat models that answer directly. Its modest size means it runs comfortably on modest consumer GPUs, making it accessible for local experimentation with reasoning-style outputs without specialized hardware. It offers a 128K token context window, ample for most reasoning and coding tasks, and is released under the MIT license, allowing broad local and commercial use.

Qwen2.5 3B Instruct

Alibaba · 3.1B · runs from 1.4 GB

Qwen2.5 3B Instruct is a 3.1-billion parameter instruction-tuned model from Alibaba Cloud's Qwen 2.5 family. It is designed for efficient local inference on consumer hardware, supporting a 128K token context window despite its compact footprint. The model can run on GPUs with as little as 4GB of VRAM when quantized. Despite its small size, Qwen2.5 3B Instruct delivers competitive performance for basic conversational tasks, summarization, and simple instruction following. It is a good option for edge deployment and resource-constrained environments. Released under the Apache 2.0 license.

GPT OSS 20B

OpenAI · 20.9B · runs from 9.3 GB

GPT-OSS 20B is one of OpenAI's first open-source model releases, marking a historic shift in the company's approach to open weights. At 21.5 billion parameters it delivers strong general-purpose chat and reasoning capabilities informed by the research behind the GPT family, making it a compelling option for users who want OpenAI-grade quality in a locally deployable package. The model runs comfortably on a single high-end consumer GPU such as an RTX 4090 at 4-bit quantization, or on workstation cards with 24 GB or more of VRAM at higher precision. It occupies a practical middle ground between lightweight 7B models and resource-heavy 70B+ offerings.

NVIDIA Nemotron 3.5 Lightning 30B A3B BF16

NVIDIA · 31.6B · runs from 9.1 GB

NVIDIA Nemotron 3.5 Lightning 30B A3B BF16 is a mixture-of-experts chat model in NVIDIA's Nemotron line, with roughly 30 billion total parameters and about 3 billion active per token. That active-parameter count drives inference speed, while the larger total governs memory footprint, and the Lightning name points to a variant tuned for fast responses; weights are published in BF16 precision. The model offers a 256K token context window and is released under a custom license rather than a standard open-source one. With only about 3 billion parameters active per token, it is well suited to running on a single high-end consumer GPU, especially once quantized, despite its larger total parameter count.

GLM 5.3 Flash

Z.ai · 321.3B · runs from 90.2 GB

GLM 5.3 Flash is a vision-language model from Z.ai in the GLM 5 family, with roughly 321 billion parameters. The Flash designation points to a variant tuned for faster inference within the family, and the model accepts image input alongside text, making it usable for tasks that combine visual and textual understanding. It supports a 1M token context window, useful for long documents or multi-image conversations, and ships under the MIT license. At its parameter scale, local deployment needs multi-GPU or server-class hardware; most users will reach it through a hosted endpoint rather than running it on a single machine.

Gemma 4 26B A4B IT Qat Q4 0 Unquantized

Google · 26.5B · runs from 11.9 GB

Gemma 4 26B A4B IT Qat Q4 0 Unquantized is a 26.5B-parameter open language model from Google in the Gemma 4 family. It supports a context window of up to 262,144 tokens. See its VRAM requirements by quantization and which GPUs and Macs can run it locally below.

Llama 3.2 1B Instruct

Meta · 1.2B · runs from 0.4 GB

Meta Llama 3.2 1B Instruct is a 1-billion parameter instruction-tuned model from Meta, the smallest in the Llama 3.2 family. It is designed for ultra-lightweight deployment scenarios where minimal hardware resources are available, supporting a 128K token context window despite its compact size. This model is suitable for basic conversational tasks, text summarization, and simple instruction following. It can run on virtually any modern GPU and even on CPU-only setups with acceptable performance. Released under the Llama 3.2 Community License.

Qwen3 30B A3B

Alibaba · 30.5B · runs from 8.8 GB

Qwen3 30B A3B is a Mixture of Experts (MoE) model from Alibaba Cloud's Qwen 3 series, with 30 billion total parameters and approximately 3 billion active parameters per forward pass. The MoE architecture delivers quality significantly above what a standard 3B dense model could achieve, while keeping per-token compute costs low. It supports hybrid thinking mode for flexible reasoning. The model requires VRAM proportional to its full 30B parameter count for weight loading, but its low active parameter count results in fast inference throughput. It is an efficient option for users who want quality beyond dense small models without the full cost of larger architectures. Released under the Apache 2.0 license.

Qwen2.5 1.5B Instruct

Alibaba · 1.5B · runs from 0.8 GB

Qwen2.5 1.5B Instruct is a 1.5-billion parameter instruction-tuned model from Alibaba Cloud's Qwen 2.5 series. It is a lightweight model suitable for deployment on minimal hardware, including low-VRAM GPUs and even CPU-only setups with acceptable latency. It supports a 128K token context window. The model handles basic conversational tasks, simple question answering, and text generation. While limited in reasoning depth compared to larger variants, it is useful for applications where fast response times and minimal resource consumption are priorities. Released under the Apache 2.0 license.

Gemma 4 E2B IT

Google · 5.1B · runs from 2.1 GB

The smallest model discussed here from Google's Gemma 4 E-series, Gemma 4 E2B IT, comes in at roughly 5 billion parameters. It is tuned for chat and instruction-following rather than multimodal tasks, prioritizing a minimal footprint over broad input support. Models at this size are comfortable on typical consumer GPUs, and can even run on integrated graphics when quantized aggressively, making it one of the more accessible options for local deployment. It ships with a 128K token context window, matching its larger E4B sibling, and is released under the Apache 2.0 license for open commercial and research use. Google published it on March 2, 2026, alongside the E4B variant, as the entry point into the Gemma 4 generation.

GLM 4.7 Flash

Z.ai · 31.2B · runs from 9.7 GB

GLM-4.7-Flash is a Mixture-of-Experts language model from Z.ai, part of the GLM 4 family, with about 31 billion total parameters and roughly 3 billion active per token. It is a chat and agentic model, positioned by Z.ai as a lightweight option for coding, tool use, and multi-turn agent tasks, with configurable reasoning-effort settings and support for extended thinking on harder problems. It supports a 202,752-token context window and is released under the MIT license. At this size, 4-bit quantization needs roughly 18GB of memory, putting it within reach of a single high-end consumer GPU in the 16-24GB range.

Qwen3.5 2B

Alibaba · 2.3B · runs from 1.0 GB

Qwen3.5 2B is a 2.3-billion-parameter model from Alibaba's Qwen team, one of the smaller entries in the Qwen3.5 lineup (0.8B–9B) built to handle text and image input together. It uses a hybrid architecture mixing linear-attention layers with periodic full-attention layers to keep long-context inference efficient. As a vision-capable model it can read and reason about images alongside written prompts, and its small size lets it run on a laptop or even a phone once quantized, without a dedicated GPU. It supports an unusually long 262K token context window for its size. It is released under the Apache 2.0 license, allowing unrestricted commercial and research use, and was published in late February 2026 as part of Alibaba's push toward compact, natively multimodal edge models.

Gemma 4 E4B IT Qat Q4 0 Unquantized

Google · 7.9B · runs from 3.9 GB

Gemma 4 E4B IT Qat Q4 0 Unquantized is a 7.9B-parameter open language model from Google in the Gemma 4 family. It supports a context window of up to 131,072 tokens. See its VRAM requirements by quantization and which GPUs and Macs can run it locally below.

Inkling

Thinking Machines · 952.4B · runs from 405.5 GB

Inkling is the flagship model in Thinking Machines' Inkling family, a vision-language chat model with roughly 952 billion parameters. It accepts image input alongside text and supports general conversational use, and was released about two weeks ahead of the smaller Inkling Small model in the same line, in mid-July 2026. Inkling is distributed under the Apache 2.0 license, allowing free local and commercial use without restriction. Given its scale, it needs multi-GPU or server-class hardware to run at all; most users will access it through a hosted endpoint rather than running it on their own machines.

MiniCPM SALA

OpenBMB · 9.5B · runs from 19.3 GB

MiniCPM SALA is a 9.5B-parameter open language model from OpenBMB in the MiniCPM family. It supports a context window of up to 524,288 tokens. See its VRAM requirements by quantization and which GPUs and Macs can run it locally below.

LFM2.5 8B A1B

Liquid AI · 8.5B · runs from 4 GB

LFM2.5-8B-A1B is Liquid AI's Mixture-of-Experts model for on-device deployment, with 8.3 billion total parameters but only about 1.5 billion active per token, built on the hybrid LFM2 architecture that mixes gated convolution blocks with grouped-query attention. It is reasoning-tuned for on-device personal-assistant use, chaining tool calls and following complex instructions, though it is not the best fit for heavy programming or knowledge-intensive question answering without retrieval. It has a 128K token context window and is released under Liquid AI's custom LFM1.0 license. Because only 1.5 billion parameters activate per token it runs quickly, though the full 8.3 billion still need to be held in memory, roughly 5 GB at 4-bit, well within reach of almost any modern laptop or GPU.

Qwen2.5 72B Instruct

Alibaba · 72.7B · runs from 21.0 GB

Qwen2.5 72B Instruct is the flagship model of the Qwen 2.5 series from Alibaba Cloud, with 72.7 billion parameters. It is instruction-tuned for conversational use and excels across reasoning, coding, mathematics, and multilingual tasks. Qwen2.5 72B delivers performance competitive with leading open-weight 70B-class models while supporting a 128K token context window and structured output generation. The model uses a Transformer architecture with grouped-query attention and was pretrained on a diverse multilingual corpus of over 18 trillion tokens. Running it locally requires high-VRAM GPUs or multi-GPU setups, though quantized formats make it accessible on workstation-class hardware. Released under the Apache 2.0 license.

Qwen3.5 0.8B

Alibaba · 873M · runs from 0.6 GB

Qwen3.5-0.8B is the smallest model in Alibaba's Qwen3.5 generation, a dense sub-billion-parameter vision-language model using the same hybrid Gated DeltaNet and gated-attention design as the rest of the family. It is a post-trained, instruction-tuned release, and its publisher notes that at this scale it is best suited to prototyping, task-specific fine-tuning, and other research or development purposes rather than production chat. It supports a 262,144-token native context window and is released under the Apache 2.0 license. At under one billion parameters, it needs well under a gigabyte of memory at 4-bit, so it runs on essentially any modern laptop or GPU.

LFM2.5 230M

Liquid AI · 230M · runs from 0.5 GB

LFM2.5-230M is Liquid AI's smallest model in the LFM2.5 family, a 230-million-parameter checkpoint distilled from the larger LFM2.5-350M and refined with multi-stage reinforcement learning for lightweight agentic tasks. It is instruction-tuned for data extraction and simple tool-use pipelines under the tightest memory and compute budgets, and the publisher does not recommend it for reasoning-heavy work like advanced math, code generation, or creative writing. It is released under Liquid AI's custom LFM1.0 license. At 230 million parameters, it runs on essentially any modern device, including phones and low-power boards like a Raspberry Pi, without needing a dedicated GPU at all.

Gemma 4 31B IT Qat Q4 0 Unquantized

Google · 32.7B · runs from 15.5 GB

Gemma 4 31B IT Qat Q4 0 Unquantized is a 32.7B-parameter open language model from Google in the Gemma 4 family. It supports a context window of up to 262,144 tokens. See its VRAM requirements by quantization and which GPUs and Macs can run it locally below.

Qwen2.5 VL 7B Instruct

Alibaba · 8.3B · runs from 17 GB

Qwen2.5 VL 7B Instruct is Alibaba's 8.3-billion-parameter vision-language model in the Qwen 2.5 lineup, built to handle images and video alongside text in a single pass. It can read documents, charts, and screenshots, describe and reason about visual content, and point out object locations within an image, useful for document understanding, visual question answering, and lightweight visual-agent tasks. At this parameter count, local inference is practical on a single mainstream or high-end consumer GPU once the weights are quantized. The model supports a 128K token context window, enough for long documents or extended multi-turn visual conversations. It is released under the Apache 2.0 license, allowing unrestricted commercial and research use, and was published in January 2025 as part of Qwen's second-generation vision-language line, later succeeded by Qwen3-VL.

GLM 5.3

Z.ai · 753.3B · runs from 258.5 GB

GLM 5.3 continues Z.ai's GLM 5 line as a mixture-of-experts model with roughly 753 billion total parameters. Like other MoE designs, it activates only a fraction of those parameters for each token, which keeps inference more efficient relative to a dense model of comparable total size. The model supports a 1M token context window for very long documents or sessions, and is distributed under a custom license rather than the MIT terms used elsewhere in the family. Given its total parameter count, GLM 5.3 needs multi-GPU or server-class hardware to run; most users will access it through a hosted endpoint instead of locally.

Qwen3 Coder Next

Alibaba · 79.7B · runs from 22.3 GB

Qwen3 Coder Next is a 79.7-billion parameter code-specialized instruction-tuned model from Alibaba Cloud, the next generation of the Qwen Coder series. It is trained extensively on source code and programming-related data, delivering strong performance across code generation, completion, debugging, refactoring, and software engineering dialogue. The model represents a significant step up in coding capability within the Qwen family. Due to its large parameter count, running Qwen3 Coder Next locally requires substantial VRAM, typically 48GB or more at reduced precision, placing it in the territory of professional GPUs or multi-GPU consumer setups. It is a top-tier choice for developers who need the most capable local coding assistant available. Released under the Apache 2.0 license.

LFM2.5 1.2B Instruct

Liquid AI · 1.2B · runs from 0.9 GB

LFM2.5 1.2B Instruct is an instruction-tuned model from Liquid AI that uses a novel hybrid architecture combining state-space models with attention mechanisms. At just 1.2 billion parameters, it is exceptionally lightweight and can run on virtually any hardware, including laptops and edge devices. Liquid AI's unconventional architecture aims to deliver better efficiency and longer context handling than traditional transformer models at this scale, making it an interesting option for users exploring alternatives to standard transformer-based LLMs.

Gemma 3 1B IT

Google · 1000M · runs from 0.3 GB

Google Gemma 3 1B IT is a 1-billion parameter instruction-tuned model from Google's Gemma 3 family. It is an ultra-compact text-only chat model designed for deployment on minimal hardware, including low-VRAM GPUs and edge devices. The model handles basic conversational tasks, simple instruction following, and lightweight text generation. It can run on virtually any modern GPU and even on CPU-only setups with acceptable latency. Released under the Gemma license.

Ornith 1.0 9B

Deep Reinforce · 9.4B · runs from 3.7 GB

Ornith 1.0 9B is Deep Reinforce's smaller, dense sibling to Ornith 1.0 35B, with roughly 9.4 billion parameters and no mixture-of-experts routing. It is tuned for general chat and instruction-following use. Models in this parameter range run comfortably on modest consumer GPUs, and remain usable on lower-VRAM setups at higher quantization, making it a practical option for local deployment without high-end hardware. The model supports a 256K token context window, matching its larger sibling despite the smaller parameter count. It is released under the MIT license and was published on June 21, 2026, giving Deep Reinforce two size options in the Ornith 1.0 line for different hardware budgets.