Model Gallery

Discover and install AI models from our curated collection

16 models available
1 repositories
Documentation

Find Your Perfect Model

Filter by Model Type

Browse by Tags

qwen3-4b-dflash
Qwen3-4B paired with its DFlash block-diffusion drafter for speculative decoding on the llama.cpp backend. This is the canonical DFlash pairing documented upstream (`z-lab/Qwen3-4B-DFlash` + `Qwen/Qwen3-4B`). DFlash produces a whole block of draft tokens in a single forward pass and injects the target model's hidden states into the drafter's attention, which keeps the drafter tiny while making drafting GPU-friendly. The Q4_K_M file carries the full Qwen3-4B target; the ~0.5 GB Q8_0 drafter (`draft-dflash`) accelerates generation without changing the target's outputs. The drafter is not a standalone chat model: it only runs paired with the target, which is why both are bundled here. Flash attention is required for DFlash and is enabled in this config. A GPU is recommended. License: Apache 2.0 (Qwen3-4B target) / MIT (z-lab DFlash drafter).

Repository: localaiLicense: apache-2.0

qwen3.5-4b-dflash
Qwen3.5-4B paired with its DFlash block-diffusion drafter for speculative decoding on the llama.cpp backend. DFlash produces a whole block of draft tokens in a single forward pass and injects the target model's hidden states into the drafter's attention, which keeps the drafter tiny while making drafting GPU-friendly. The Q4_K_M file carries the full Qwen3.5-4B target; the ~0.6 GB Q8_0 drafter (`draft-dflash`) accelerates generation without changing the target's outputs. The drafter is not a standalone chat model: it only runs paired with the target, which is why both are bundled here. Flash attention is required for DFlash and is enabled in this config. A GPU is recommended. License: Apache 2.0 (Qwen3.5-4B target) / MIT (z-lab DFlash drafter).

Repository: localaiLicense: apache-2.0

qwen3.5-9b-dflash
Qwen3.5-9B paired with its DFlash block-diffusion drafter for speculative decoding on the llama.cpp backend. DFlash produces a whole block of draft tokens in a single forward pass and injects the target model's hidden states into the drafter's attention, which keeps the drafter tiny while making drafting GPU-friendly. The Q4_K_M file carries the full Qwen3.5-9B target; the ~1 GB Q8_0 drafter (`draft-dflash`) accelerates generation without changing the target's outputs. The drafter is not a standalone chat model: it only runs paired with the target, which is why both are bundled here. Flash attention is required for DFlash and is enabled in this config. A GPU is recommended. License: Apache 2.0 (Qwen3.5-9B target) / MIT (z-lab DFlash drafter).

Repository: localaiLicense: apache-2.0

qwen3.6-27b-dflash
Qwen3.6-27B (dense) paired with its DFlash block-diffusion drafter for speculative decoding on the llama.cpp backend. DFlash gives its largest speedups on dense targets like this one. DFlash produces a whole block of draft tokens in a single forward pass and injects the target model's hidden states into the drafter's attention, which keeps the drafter tiny while making drafting GPU-friendly. The Q4_K_M file carries the full Qwen3.6-27B target; the ~1.8 GB Q8_0 drafter (`draft-dflash`) accelerates generation without changing the target's outputs. The drafter is not a standalone chat model: it only runs paired with the target, which is why both are bundled here. Flash attention is required for DFlash and is enabled in this config. A GPU is recommended. License: Apache 2.0.

Repository: localaiLicense: apache-2.0

qwen3.6-35b-a3b-dflash
Qwen3.6-35B-A3B (Mixture-of-Experts, ~3B active per token) paired with its DFlash block-diffusion drafter for speculative decoding on the llama.cpp backend. DFlash speedups on MoE targets are smaller than on dense models, but still useful. DFlash produces a whole block of draft tokens in a single forward pass and injects the target model's hidden states into the drafter's attention, which keeps the drafter tiny while making drafting GPU-friendly. The UD-Q4_K_M file carries the full Qwen3.6-35B-A3B target; the ~0.4 GB Q8_0 drafter (`draft-dflash`) accelerates generation without changing the target's outputs. The drafter is not a standalone chat model: it only runs paired with the target, which is why both are bundled here. Flash attention is required for DFlash and is enabled in this config. A GPU is recommended. License: Apache 2.0.

Repository: localaiLicense: apache-2.0

glm-5.2
# GLM-5.2 👋 Join our WeChat or Discord community. 📖 Check out the GLM-5.2 blog and GLM-5 Technical report. 📍 Use GLM-5.2 API services on Z.ai API Platform. 🔜 Try GLM-5.2 here. [Paper] [GitHub] ## Introduction We're introducing GLM-5.2, our latest flagship model for long-horizon tasks. It marks a substantial leap in long-horizon task capability over its predecessor GLM-5.1 and, for the first time, delivers that capability on a **solid 1M-token context**. GLM-5.2's new capabilities include: - **Solid 1M Context:** A solid 1M-token context that stably sustains long-horizon work - **Advanced Coding with Flexible Effort**: Stronger coding capabilities with multiple thinking effort levels to balance performance and latency - **Improved Architecture**: We propose IndexShare, which reuses the same indexer across every four sparse attention layers, reducing per-token FLOPs by 2.9× at a 1M context length. We also improve GLM-5.2’s MTP layer for speculative decoding, increasing the acceptance length by up to 20% - **Pure Open**: An MIT open-source license — no regional limits, technical access without borders ## Benchmark ## Serve GLM-5.2 Locally ...

Repository: localaiLicense: mit

gemma-4-e2b-it-qat-mtp
Gemma 4 E2B IT QAT (Google DeepMind) paired with its Multi-Token Prediction (MTP) drafter head for speculative decoding on the llama.cpp backend. The Q4_K_XL target carries the full multimodal (text + image) model; the small `mtp-gemma-4-E2B-it` head predicts several tokens ahead which the target verifies in parallel, accelerating generation with no change to output quality. E2B is a MatFormer "effective 2B" elastic variant, well suited to lightweight and on-device deployments. The drafter is not a standalone chat model: it only runs paired with the target, which is why both are bundled here. It uses the upstream `gemma4-assistant` architecture registered by llama.cpp PR #23398, so it loads on stock llama.cpp without any patch. License: Apache 2.0 | Authors: Google DeepMind (target/drafter checkpoints), Unsloth (GGUF conversion)

Repository: localaiLicense: apache-2.0

gemma-4-e4b-it-qat-mtp
Gemma 4 E4B IT QAT (Google DeepMind) paired with its Multi-Token Prediction (MTP) drafter head for speculative decoding on the llama.cpp backend. The Q4_K_XL target carries the full multimodal (text + image) model; the small `mtp-gemma-4-E4B-it` head predicts several tokens ahead which the target verifies in parallel, accelerating generation with no change to output quality. E4B is a MatFormer "effective 4B" elastic variant, balancing quality and footprint for on-device and edge deployments. The drafter is not a standalone chat model: it only runs paired with the target, which is why both are bundled here. It uses the upstream `gemma4-assistant` architecture registered by llama.cpp PR #23398, so it loads on stock llama.cpp without any patch. License: Apache 2.0 | Authors: Google DeepMind (target/drafter checkpoints), Unsloth (GGUF conversion)

Repository: localaiLicense: apache-2.0

gemma-4-12b-it-qat-mtp
Gemma 4 12B IT QAT (Google DeepMind) paired with its Multi-Token Prediction (MTP) drafter head for speculative decoding on the llama.cpp backend. The Q4_K_XL target carries the full multimodal (text + image) model; the small `mtp-gemma-4-12B-it` head predicts several tokens ahead which the target verifies in parallel, accelerating generation with no change to output quality. As a dense model, Gemma 4 12B is among the sizes that benefit most from MTP, with the llama.cpp PR reporting well over 1.4x decode speedup. The drafter is not a standalone chat model: it only runs paired with the target, which is why both are bundled here. It uses the upstream `gemma4-assistant` architecture registered by llama.cpp PR #23398, so it loads on stock llama.cpp without any patch. License: Apache 2.0 | Authors: Google DeepMind (target/drafter checkpoints), Unsloth (GGUF conversion)

Repository: localaiLicense: apache-2.0

gemma-4-31b-it-qat-mtp
Gemma 4 31B IT QAT (Google DeepMind), the largest dense model in the family, paired with its Multi-Token Prediction (MTP) drafter head for speculative decoding on the llama.cpp backend. The Q4_K_XL target carries the full multimodal (text + image) model; the small `mtp-gemma-4-31B-it` head predicts several tokens ahead which the target verifies in parallel, accelerating generation with no change to output quality. Dense models like 31B are the sizes that benefit most from MTP. The drafter is not a standalone chat model: it only runs paired with the target, which is why both are bundled here. It uses the upstream `gemma4-assistant` architecture registered by llama.cpp PR #23398, so it loads on stock llama.cpp without any patch. License: Apache 2.0 | Authors: Google DeepMind (target/drafter checkpoints), Unsloth (GGUF conversion)

Repository: localaiLicense: apache-2.0

alamios_mistral-small-3.1-draft-0.5b
This model is meant to be used as draft model for speculative decoding with mistralai/Mistral-Small-3.1-24B-Instruct-2503 or mistralai/Mistral-Small-24B-Instruct-2501 Data info The data are Mistral's outputs and includes all kind of tasks from various datasets in English, French, German, Spanish, Italian and Portuguese. It has been trained for 2 epochs on 20k unique examples, for a total of 12 million tokens per epoch.

Repository: localaiLicense: apache-2.0

gemma-4-e2b-it:sglang-mtp
Google Gemma 4 E2B-IT served by SGLang with Multi-Token Prediction (MTP) speculative decoding. The companion drafter google/gemma-4-E2B-it-assistant lets the target accept several tokens per step. Flags are a 1:1 transcription of the SGLang cookbook's MTP command (NEXTN algorithm, num_steps=5, num_draft_tokens=6, eagle_topk=1, mem_fraction_static=0.85). The E2B variant has 5B total / 2B effective parameters and targets the smaller end of consumer GPUs.

Repository: localaiLicense: gemma

gemma-4-e4b-it:sglang-mtp
Google Gemma 4 E4B-IT served by SGLang with Multi-Token Prediction (MTP) speculative decoding. The companion drafter google/gemma-4-E4B-it-assistant lets the target accept several tokens per step. Flags are a 1:1 transcription of the SGLang cookbook's MTP command (NEXTN algorithm, num_steps=5, num_draft_tokens=6, eagle_topk=1, mem_fraction_static=0.85). The E4B variant has 8B total / 4B effective parameters — the natural pick for consumer GPUs in the 16–24 GB range.

Repository: localaiLicense: gemma

mimo-7b-mtp:sglang
Xiaomi MiMo-7B-RL served by SGLang with built-in Multi-Token Prediction (MTP) heads (no separate drafter needed) plus online fp8 weight quantization to fit on a 16 GB consumer GPU. ~90% acceptance per the model card. Verified end-to-end at ~88 tok/s on an RTX 5070 Ti (16 GB). Note: mem_fraction_static is dropped to 0.7 (vs sglang's 0.85 default) because the MTP draft worker's vocab embedding is loaded unquantised (~1.2 GiB) and OOMs the static reservation otherwise.

Repository: localaiLicense: mit

almost-human-x3-32bit-1839-6b-i1
**Model Name:** Almost-Human-X3-32bit-1839-6B **Base Model:** Qwen3-Jan-v1-256k-ctx-6B-Brainstorm20x **Author:** DavidAU **Repository:** [DavidAU/Almost-Human-X3-32bit-1839-6B](https://huggingface.co/DavidAU/Almost-Human-X3-32bit-1839-6B) **License:** Apache 2.0 --- ### 🔍 **Overview** A high-precision, full-precision (float32) fine-tuned variant of the Qwen3-Jan model, specifically trained to emulate the literary and philosophical depth of Philip K. Dick. This model is the third in the "Almost-Human" series, built with advanced **"Brainstorm 20x"** methodology to enhance reasoning, coherence, and narrative quality—without sacrificing instruction-following ability. ### 🎯 **Key Features** - **Full Precision (32-bit):** Trained at 16-bit for 3 epochs, then finalized at float32 for maximum fidelity and performance. - **Extended Context (256k tokens):** Ideal for long-form writing, complex reasoning, and detailed code generation. - **Advanced Reasoning via Brainstorm 20x:** The model’s reasoning centers are expanded, calibrated, and interconnected 20 times, resulting in: - Richer, more nuanced prose - Stronger emotional engagement - Deeper narrative focus and foreshadowing - Fewer clichés, more originality - Enhanced coherence and detail - **Optimized for Creativity & Code:** Excels at brainstorming, roleplay, storytelling, and multi-step coding tasks. ### 🛠️ **Usage Tips** - Use **CHATML or Jinja templates** for best results. - Recommended settings: Temperature 0.3–0.7 (higher for creativity), Top-p 0.8, Repetition penalty 1.05–1.1. - Best used with **"smoothing" (1.5)** in GUIs like KoboldCpp or oobabooga. - For complex tasks, use **Q6 or Q8 GGUF quantizations**. ### 📦 **Model Formats** - **Full precision (safe tensors)** – for training or high-fidelity inference - **GGUF, GPTQ, EXL2, AWQ, HQQ** – available via quantization (see [mradermacher/Almost-Human-X3-32bit-1839-6B-i1-GGUF](https://huggingface.co/mradermacher/Almost-Human-X3-32bit-1839-6B-i1-GGUF) for quantized versions) --- ### 💬 **Ideal For** - Creative writing, speculative fiction, and philosophical storytelling - Complex code generation with deep reasoning - Roleplay, character-driven dialogue, and immersive narratives - Researchers and developers seeking a highly expressive, human-like model > 📌 **Note:** This is the original source model. The GGUF versions by mradermacher are quantized derivatives — not the base model. --- **Explore the source:** [DavidAU/Almost-Human-X3-32bit-1839-6B](https://huggingface.co/DavidAU/Almost-Human-X3-32bit-1839-6B) **Quantization guide:** [mradermacher/Almost-Human-X3-32bit-1839-6B-i1-GGUF](https://huggingface.co/mradermacher/Almost-Human-X3-32bit-1839-6B-i1-GGUF)

Repository: localaiLicense: apache-2.0

deepseek-v4-flash-q2-mtp
DeepSeek V4 Flash (IQ2XXS GGUF, ~81 GB) paired with the optional MTP speculative-decoding weights (~3.5 GB) for a slight speedup. Only loadable via the ds4 backend; requires >=128 GB RAM. MTP helps only with greedy decoding (temperature 0), so the override pins temperature to 0. Metal (Darwin) or CUDA (Linux). See https://github.com/antirez/ds4 for details.

Repository: localai