CatalogWantedMirrorsLedgerOperating StatementDocsStudio statusCard system
alt.freeweights.text / KAT-Coder-V2.5-Dev-OptiQ-4bit
From: mlx-community <huggingface.co/mlx-community>Newsgroups: alt.freeweights.textSubject:

KAT-Coder-V2.5-Dev-OptiQ-4bit

Date: 2026-11-15Commit: 57c6e83Signed: 2DC28CC7B26A283B ✓

Text generation model with 34.7B parameters. Safetensors weights, 22.0 GB in all. Published by mlx-community under Apache-2.0.

34.7B params22.0GBKAT-Coder-V2.5-Dev-OptiQ-4bitsafetensorspermissive
FactsParams34.7BSize22.0 GBFormatssafetensorsTasktext generationLicenseApache-2.0From sourcepermissive
$ fw get mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit
# manifest sha256 92446b9a32a0… · 22.0 GB · 15 files
$ fw verify ./KAT-Coder-V2.5-Dev-OptiQ-4bit mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit
From sourceHosted (signed)mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bitLast check: 2026-11-15 04:40 UTC
Run at homeGet help

Get help: custom fine-tune or private deploy. We review your setup before we quote.

Quantized of KAT-Coder-V2.5-Dev.

Works with

Run24Fine-tune29Convert15Evaluate13Scan13Apps58
OverviewModel cardFilesUse itLicenseProvenanceModel tree

Overview

Size
22.0 GB
Weights
21.9 GB (safetensors)
Parameters
34.7B
Formats
safetensors
Task
text generation
License
Apache-2.0 (permissive)
Files
15, each with a sha256
Publisher
mlx-community
Added
2026-11-15
Section
alt.freeweights.text
Lineage

Original release from the publisher.

Base model: Kwaipilot/KAT-Coder-V2.5-Dev

Where these bytes come from
Publisher at the pinned commit→You

Sources are tried in order. A file counts only when its size and sha256 match the manifest. Sources are checked when you ask for them, not on a timer.

Sources
BadgeHostLast checkResult
From sourcehuggingface.co2026-11-15 04:40 UTCpassed

Model card

From the publisher at commit 57c6e83. Shown as plain text with links; images open at the publisher.

KAT-Coder-V2.5-Dev-OptiQ-4bit#

Built with mlx-optiq, the MLX-native toolkit to quantize, fine-tune, and serve LLMs locally on Apple Silicon (no PyTorch, no cloud). Try the Lab · All OptiQ quants · Docs

An OptiQ mixed-precision MLX quant of KAT-Coder-V2.5-Dev, a qwen3_5_moe coding model (256-routed-expert sparse MoE with hybrid linear + full attention).

  • Mixed 4/8-bit static build — per-layer bit-widths assigned to a 4.5 target bits-per-weight: 400 projections at 4-bit, 111 at 8-bit.
  • ~4.51 bits per weight, 20 GB on disk.

Requirements#

pip install -U optiq

qwen3_5_moe loads under stock mlx-lm too, but optiq serve adds mixed- precision loading, KV-cache quantization, and the OptiQ Lab.

Running it#

optiq serve --model mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit

Then use the OpenAI-compatible endpoint at http://localhost:8000/v1, the OptiQ Lab, or point optiq code at it. This is a reasoning coder — it thinks before it answers.

Files

15 files · 22.0 GBWeights 21.9 GB (safetensors). Sizes and sha256 from the signed manifest.
Hide all 15 files with their sha256
FileSizesha256
.gitattributes1.57 KB34448b82c17d60fec9b65b1f093c115ddbaadc04beb1b0140b6bfed2e012a930
README.md1.37 KBf01bdde47530dcfb33e50b1d4d0f33954a19bb5fb4c83cfe7eb470cff27fddab
chat_template.jinja7.76 KBe84f32a23fdda27689f868aa4a1a5621f41133e51a48d7f3efcbea2839574259
config.json115 KBc2c5be70d6141247b836394a4f6a0b7ffa2860f84155ebbf78a8669b438072db
generation_config.json202 Be70c136c1b78ddc1fb0905bac8e733a4dc448d4f852a5dd75143fffc70be550e
kv_config.json660 Bc91b198188237b8de5588f4cc607391618fc6159a17f6035b880ec18fee7e6c9
model-00001-of-00005.safetensors5.31 GBf0b8ab10f7c280f408086ed9a97ce9adfb147ca484f2e963d3773a508e61f5a4
model-00002-of-00005.safetensors5.37 GB01b42106ea428c947a34379e81a2f51fcb2454b7e945f4a0f60c785521589cd2
model-00003-of-00005.safetensors5.25 GBac181b1330eb97a3550854b957f1a651b61ceda7bdd6ad4c726455311855d32f
model-00004-of-00005.safetensors5.18 GB4e2b85371b711a8227fe314850b7c7d4efad1cd2b610d64facf3747a1365f93c
model-00005-of-00005.safetensors829 MB5a55d7784c46d7e3deaf7fbf1ea4bdfa814f5dae22e40c0379c7c5114a2a34ab
model.safetensors.index.json187 KB6257a64d9ca5c68348ff55944f8db6d40856cc3a5d73ff379f80d5849116306c
optiq_metadata.json56.2 KBc3fa0e4d0e1f8057501bb993c0235bf37276d9f0c33ebc40d896a218ba83610c
tokenizer.json20.0 MB06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
tokenizer_config.json1.16 KB95c557768e6b88a7128befc7bfd3c7de50e5d51af9b8b33a9f4dee0e04f99679
VersionsEach time the pinned upstream commit or file list changed in this repo. Newest first.
Upstream commitDateContents
57c6e83727 pinned now2026-11-15 04:40 UTC15 files · 22.0 GB

Use it

$ fw get mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit
# manifest sha256 92446b9a32a0… · 22.0 GB · 15 files
$ fw verify ./KAT-Coder-V2.5-Dev-OptiQ-4bit mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit
fw.pyChecks the signature and every sha256, tries each source in order, resumes.
$ python3 fw.py get https://freeweights.allternit.com/manifests/mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit.json --dir ./KAT-Coder-V2.5-Dev-OptiQ-4bit $ python3 fw.py verify ./KAT-Coder-V2.5-Dev-OptiQ-4bit ./mlx-community__KAT-Coder-V2.5-Dev-OptiQ-4bit.json
hfThe Hugging Face CLI with one setting. Same repo id.
$ HF_ENDPOINT=https://freeweights.allternit.com hf download mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit --revision 57c6e8372737f8a04aca1c210c926b8aa29c802e --local-dir ./KAT-Coder-V2.5-Dev-OptiQ-4bit
transformersLoad the weights in Python.
import os os.environ["HF_ENDPOINT"] = "https://freeweights.allternit.com" from transformers import AutoModelForCausalLM, AutoTokenizer tok = AutoTokenizer.from_pretrained("mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit", revision="57c6e8372737f8a04aca1c210c926b8aa29c802e") model = AutoModelForCausalLM.from_pretrained("mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit", revision="57c6e8372737f8a04aca1c210c926b8aa29c802e", device_map="auto")
vLLMServe an OpenAI-compatible API.
$ HF_ENDPOINT=https://freeweights.allternit.com vllm serve mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit --revision 57c6e8372737f8a04aca1c210c926b8aa29c802e
Run at home

License

Name
Apache License 2.0
Id
Apache-2.0
Class
Permissive
Redistribution
yes
Listing
Hosted (signed)
Decided
License as declared by the publisher. No person read it. 2026-11-15
Text in
README.md
Text sha256
f01bdde47530dcfb33e50b1d4d0f33954a19bb5fb4c83cfe7eb470cff27fddab
Terms gate
No extra terms beyond the license.

Read the license text

Licenses are checked against the license table. A custom license passes only after a person reads it.

Provenance

From: mlx-community <huggingface.co/mlx-community>Newsgroups: alt.freeweights.textSubject: KAT-Coder-V2.5-Dev-OptiQ-4bitDate: 2026-11-15Commit: 57c6e83Signed: 2DC28CC7B26A283B ✓
Upstream
mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit
Pinned commit
57c6e8372737f8a04aca1c210c926b8aa29c802e
Manifest
KAT-Coder-V2.5-Dev-OptiQ-4bit.json · .minisig
Manifest sha256
92446b9a32a0560682a1acbd598f561779865b3655a9c934432639b6b84e6822
Signature
minisign, Ed25519, key 2DC28CC7B26A283B
Created
2026-11-15 04:40 UTC
Generator
freeweights-manifest/0.1.0
Takedowns
None.

How to verify

$ curl -O https://freeweights.allternit.com/manifests/mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit.json $ curl -O https://freeweights.allternit.com/manifests/mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit.json.minisig $ minisign -Vm KAT-Coder-V2.5-Dev-OptiQ-4bit.json -P RWQ7KGqyx4zCLXtUFOqvfv8L3g+8fpKJ1FsuxPyOYMF+4KufbUxKq1Oq $ shasum -a 256 <file> # compare with the Files tab

A manifest whose signature fails must be rejected. A file whose size or sha256 does not match must be discarded. Verify a download

Model tree

Parent

Kwaipilot/KAT-Coder-V2.5-Dev · this model is a quantization of it

1 model in the catalog name this one as their parent.

Quantizations
1
Quantizations 1
Get it
Download
$ HF_ENDPOINT=https://freeweights.allternit.com hf download mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit
Open in Freeweights Studio

Pull, verify and run it in the Allternit app on Desktop, web or phone.

Signed manifest
minisign, key 2DC28CC7B26A283B. KAT-Coder-V2.5-Dev-OptiQ-4bit.json. Provenance. Takedown history: none.
Source
huggingface.co/mlx-community/KAT-Coder-V2.5-Dev-OptiQ-4bit

Variants

1 quantized or adapter build of KAT-Coder-V2.5-Dev-OptiQ-4bit. Each has its own license check and page.

SubjectSizeParamsLicenseDate
KAT-Coder-V2.5-Dev-OptiQ-4bit-REAP-18BHosted (signed) From: mlx-community in alt.freeweights.text12.3 GB18.5BPermissive2026-11-15

Related models

All of alt.freeweights.text →
From: laion <huggingface.co/laion>Subject: a2-rl-e2egit_large-50-32B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_exp_rpt_e2egit-large-15-8B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_exp_rpt_e2egit-v2-10-8B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_exp_rpt_pymethods2test-large-80-8B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_exp_rpt_unitsyn-python-large-10-8B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_mix_h2_language_proportional-65-8B