Model Gallery
Discover and install AI models from our curated collection
Find Your Perfect Model
Filter by Model Type
Browse by Tags
deepseek-v4.1-flash-ista-iq1_m
DeepSeek-V4.1-Flash-1017, a 763B-parameter mixture-of-experts model (384 experts
per layer, 6 used per token), in ISTA-DAS's 1.75 bits per weight GSQ-RCO
quantization (IQ1_M), the smallest size of the three (about 239 GB to download).
It writes a reasoning trace before the answer, and the Strata backend splits it
from the reply and parses the model's DSML tool calls. Text only: images are not
supported by this backend.
Needs a GPU with 12 GB of VRAM or more and about 34 GB of RAM at the least
(Strata's figure; its measurements use 46 to 64 GB: an RTX 5070 with 64 GB
answered at about 5 tokens per second, an RTX 3060 with 46 GB at 8 to 10). Less
RAM still runs, slowly. Slightly lower quality than 2.00 bpw, and noticeably
weaker on plain English prose; ISTA-DASLab measured 98.7% of the base model on
its task average (AIME25, GPQA-Diamond, LiveCodeBench v6).
Needs about 240 GB of free disk, on a fast NVMe drive: the 111 GB n-gram table
(Engram) stays on disk and is read for every token. The Engram table and the 8
GB DSpark draft model are shared by all DeepSeek-V4.1 entries and downloaded
once into strata/models/deepseek-v4.1-flash-shared.
Runs only on the experimental strata backend, on NVIDIA CUDA 13 (x86_64 and
arm64), AMD Radeon RX 9070 and R9700 (ROCm) and Apple Silicon (Metal). Strata
ships no DeepSeek engine for CUDA 12, Intel or other AMD cards.
On the first load the strata backend builds the Strata pack from the GGUF files
(under a minute, a few MB of extra disk); loading the model afterwards takes a
few minutes. The default context is 8192 tokens, the size Strata checked; raise
context_size if the GPU has room.
Licensed under the MIT License (DeepSeek-V4.1-Flash-1017 and the quantized
files). The GGUF files are the GSQ-RCO quantization by ISTA-DAS (ISTA-DASLab);
the DSpark draft model is Unsloth's; Strata is by Niko1221.
Repository: localaiLicense: mit
qwen3.8-flash-next-nvfp4-nvidia
Qwen3.8-Flash-Next, a 125B-parameter hybrid mixture-of-experts model, in
NVIDIA's NVFP4 quantization (nvidia/Qwen3.8-Flash-Next-NVFP4), converted to a
Strata pack with GGUF side files (about 129 GB to download). It reads images
through the vision projector published by ISTA-DASLab for the original model,
and the Strata backend parses its reasoning and tool calls.
Needs an NVIDIA GPU with 12 GB of VRAM or more (RTX 20 series or newer, or a
DGX Spark) and about 80 GB of RAM (about 69 GiB of expert weights are kept in
memory), plus fast SSD storage for the 48 GiB n-gram table. Runs only on the
NVIDIA builds of the experimental strata backend.
Two licences apply: the NVIDIA Open Model License Agreement (NVIDIA's
quantization) and the Qwen Community License 1.0 (the base model). The
conversion to the Strata format is by Maximilian228. The MTP draft layer
shipped in the repository is not configured here.
Repository: localaiLicense: nvidia-open-model-license
sage-a1.5-q4
Sage-A1.5 is a 35B mixture-of-experts model based on Qwen3.6 for
scientific reasoning, coding, and tool use. This official Q4_K_M GGUF
build uses llama.cpp and includes the vision projector.
Repository: localaiLicense: apache-2.0
sage-a1.5-q8
Sage-A1.5 is a 35B mixture-of-experts model based on Qwen3.6 for
scientific reasoning, coding, and tool use. This official Q8_0 GGUF
build uses llama.cpp and includes the vision projector.
Repository: localaiLicense: apache-2.0
bge-base-en-v1.5-f16
BGE-base-en-v1.5 is a 109M English embedding model for semantic search
and retrieval. This F16 GGUF build produces 768-dimensional embeddings
with a 512-token context. For retrieval, prefix queries with
"Represent this sentence for searching relevant passages: "; leave passages unprefixed.
Repository: localaiLicense: mit
bge-base-en-v1.5-q8-0
BGE-base-en-v1.5 is a 109M English embedding model for semantic search
and retrieval. This Q8_0 GGUF build produces 768-dimensional embeddings
with a 512-token context. For retrieval, prefix queries with
"Represent this sentence for searching relevant passages: "; leave passages unprefixed.
Repository: localaiLicense: mit
bge-base-en-v1.5-q4-k-m
BGE-base-en-v1.5 is a 109M English embedding model for semantic search
and retrieval. This Q4_K_M GGUF build produces 768-dimensional embeddings
with a 512-token context. For retrieval, prefix queries with
"Represent this sentence for searching relevant passages: "; leave passages unprefixed.
Repository: localaiLicense: mit
ornith-1.5-9b-uncensored
# Ornith-1.5-9B-uncensored
An **abliterated** (refusal-direction-ablated) build of
`ornith-ai/Ornith-1.5-9B`, produced with
ZeroFuse and published by
junafinity.
This is the **9B control checkpoint** (bf16). Mac users should start from the MLX-8bit or GGUF-8bit siblings. The official 9B base has **no `mtp.*` tensors**; nothing was grafted.
**Vision tower and MTP heads are preserved** — see
Vision & MTP preservation for the before/after audit.
## Intended use: red teaming and defensive cybersecurity research
These uncensored (abliterated) weights are built as a **research instrument** for red teaming and defensive cybersecurity work. Safety training suppresses the *display* of capability, not capability itself. A refusal tells you the model declined. It does not tell you whether the weights could have complied. That conflation underestimates the true ceiling and hides holes in *your* filters, classifiers, and policy layer.
Use each uncensored checkpoint as the **treatment half of a controlled pair** against its original base model:
...
Repository: localaiLicense: apache-2.0
qwen3.8-flash-next-uncensored
# Qwen3.8-Flash-Next
> [!Note]
> This repository contains model weights and configuration files for the post-trained model in the Hugging Face Transformers format.
>
> These artifacts are compatible with Hugging Face Transformers, vLLM, SGLang, TokenSpeed, etc.
> [!Tip]
> For users seeking managed, scalable inference without infrastructure maintenance, the official Qwen API service is provided by Qwen Cloud.
>
> In particular, **Qwen3.8-Flash** is the official version based on Qwen3.8-Flash-Next with more production features, e.g., 1M context length by default, official built-in tools. For more information, please refer to the Qwen3.8-Flash Overview.
As the frontier of foundation models pushes toward ever-larger parameter counts and ever-longer context windows, the question is no longer just how much we can scale, but how efficiently we can do so. Sustainable progress toward artificial general intelligence (AGI) that benefits everyone demands architectural innovation. Today, we are sharing a concrete step in that direction: Qwen3.8-Flash-Next.
...
Repository: localaiLicense: apache-2.0
glm-5.3-flash
# GLM-5.3-Flash
👋 Join our WeChat or Discord community.
📖 Check out the GLM-5.3-Flash blog and GLM-5 Technical report.
📍 Use GLM-5.3-Flash API services on Z.ai API Platform.
## Introduction
We introduce GLM-5.3-Flash, the first natively multimodal model in the GLM-5 series. With 320B total parameters and just 18B active parameters, it outperforms GLM-5.2 across benchmarks and real-world workloads at one-tenth the price, while approaching Claude Opus 4.8 on coding and agentic benchmarks.
GLM-5.3-Flash starts from a newly trained base model, with its architecture and training recipe redesigned around capability and efficiency. For the first time in the GLM series, we introduce a hybrid architecture combining sparse and linear attention, sharply reducing long-context serving costs while preserving precise long-context capabilities. The model also adopts Manifold-Constrained Hyper-Connections (mHC) to further improve scaling efficiency. Together with our latest 30T-token multimodal pre-training corpus, these changes enable GLM-5.3-Flash to deliver more intelligence with less compute.
## Serve GLM-5.3-Flash Locally
...
Repository: localaiLicense: mit
glm-5.3
# GLM-5.3
GLM-5.3 uses the same base model as GLM-5.2 — every gain comes from post-training. Compared with GLM-5.2, it is much better at complex coding and long-horizon tasks:
+ Stronger Coding: GLM-5.3 is the most capable open-weights model for coding, with a 50% improvement over GLM-5.2 on our in-house Z.ai Code Bench. It also achieve open-source SOTA on public benchmarks including Terminal Bench 3.0 and Agents' Last Exam.
+ Emergent Cyber Capability: As we scaled post-training, cyber capability developed faster than we expected. GLM-5.3 is state of the art on CyberGym for vulnerability discovery, and its gains are largest further up the exploitation chain, where it more than doubles GLM-5.2 on exploitation benchmarks.
## Benchmark
### Serve GLM-5.3 Locally
GLM-5.3 supports deployment with the following frameworks. Feel free to try them out:
- SGLang — see cookbook
- vLLM — see recipes
- TokenSpeed — see here
- Transformers — see transformers docs
- KTransformers — see tutorial
- Unsloth — see guide
- For deployment on the `Ascend NPU` platform, inference frameworks such as vLLM-Ascend, xLLM and SGLang are supported — see here.
### Note
...
Repository: localaiLicense: other
pocket-darwin-180b-q4
POCKET-Darwin-180B is a text-only reasoning model based on
Qwen3.8-Flash-Next. This UD-Q4_K_XL GGUF build uses llama.cpp and
downloads all four shards (about 111.3 GB). It uses the embedded
chat template, an 8192-token default context, and the publisher's
recommended sampling settings. Weights use Qwen Community License 1.0.
Repository: localaiLicense: other
qwen3.8-27b-uncensored-q4
Qwen3.8-27B-Uncensored reduces refusal behavior while retaining the base
model's text, vision, reasoning, and tool-use capabilities. Its integrated
MTP head supports speculative decoding without a separate draft model.
This default entry uses the Q4_K_M GGUF and F16 vision projector. A
higher-quality Q8_0 model is available as a variant.
Repository: localaiLicense: apache-2.0
cyber-tiel-coder-35b-a3b-q4-mtp
Cyber-Tiel-Coder is a 35B mixture-of-experts coding model with 3B active parameters,
based on Huihui's abliterated Ornith-1.5. This UD-Q4_K_XL build includes
MTP speculative decoding, the embedded Sharp chat template, and a BF16 vision projector.
Repository: localaiLicense: mit
cyber-tiel-coder-35b-a3b-q8-mtp
Cyber-Tiel-Coder is a 35B mixture-of-experts coding model with 3B active parameters,
based on Huihui's abliterated Ornith-1.5. This UD-Q8_K_XL build includes
MTP speculative decoding, the embedded Sharp chat template, and a BF16 vision projector.
Repository: localaiLicense: mit
twil-lm3-q4
TwIL-LM3 is a 3B SmolLM3-based reasoning model specialized for formal
logic, entailment, semantic parsing, and Lean formalization. This default
entry uses the publisher's recommended Q4_K_M GGUF and supports a 65K-token
context window. A higher-quality Q8_0 build is available as a variant.
Repository: localaiLicense: webai-non-commercial-license-ver.-1.0
b1-9b
B1-9B is Schneewolf Labs' Qwen3.5-based 9B model for reasoning and tool use.
It is trained to produce an answer after reasoning about tool results.
This Q4_K_M GGUF build includes the Q8_0 vision projector and uses an
8192-token context by default. A Q8_0 variant is also available.
Repository: localaiLicense: apache-2.0
qwen3.8-35b-a3b-distill-q4
Empero's Qwen3.8 distillation into Qwen3.6-35B-A3B has 35B total
parameters and about 3B active per token. This Q4_K_M GGUF build
uses the embedded reasoning template and includes the F16 vision
projector. Vision is inherited from the base and was not evaluated
by the publisher.
Repository: localaiLicense: apache-2.0
qwen3.8-35b-a3b-distill-q5
Empero's Qwen3.8 distillation into Qwen3.6-35B-A3B has 35B total
parameters and about 3B active per token. This Q5_K_M GGUF build
uses the embedded reasoning template and includes the F16 vision
projector. Vision is inherited from the base and was not evaluated
by the publisher.
Repository: localaiLicense: apache-2.0
qwen3.8-35b-a3b-distill-q8
Empero's Qwen3.8 distillation into Qwen3.6-35B-A3B has 35B total
parameters and about 3B active per token. This Q8_0 GGUF build
uses the embedded reasoning template and includes the F16 vision
projector. Vision is inherited from the base and was not evaluated
by the publisher.
Repository: localaiLicense: apache-2.0
occamy-1.0-q4
Occamy-1.0 is a 35B MoE model with 3B active parameters, based on
Qwen3.6-35B-A3B and trained for multi-step agent tasks and coding.
This Q4_K_M GGUF build includes the F16 vision projector and uses
the embedded Jinja chat template with an 8K-token default context.
Normalize prompt text to Unicode NFC to match the source tokenizer.
Repository: localaiLicense: apache-2.0
Page
1
of many