CatalogWantedMirrorsLedgerOperating StatementDocsStudio statusCard system
alt.freeweights.text / Ling-3.1-flash-1019-oQ2.5e-mtp
From: mlx-community <huggingface.co/mlx-community>Newsgroups: alt.freeweights.textSubject:

Ling-3.1-flash-1019-oQ2.5e-mtp

Date: 2026-11-15Commit: 3f47161Signed: 2DC28CC7B26A283B ✓

Text generation model with 572.1B parameters. Safetensors weights, 224 GB in all. Published by mlx-community under MIT.

572.1B params224GBLing-3.1-flash-1019-oQ2.5e-mtpsafetensorspermissive
FactsParams572.1BSize224 GBFormatssafetensorsTasktext generationLicenseMITFrom sourcepermissive
$ fw get mlx-community/Ling-3.1-flash-1019-oQ2.5e-mtp
# manifest sha256 e68fc13f9901… · 224 GB · 52 files
$ fw verify ./Ling-3.1-flash-1019-oQ2.5e-mtp mlx-community/Ling-3.1-flash-1019-oQ2.5e-mtp
From sourceHosted (signed)mlx-community/Ling-3.1-flash-1019-oQ2.5e-mtpLast check: 2026-11-15 05:15 UTC
Fine-tune in StudioGet help

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

Quantized of Ling-3.1-flash-1019.

Runs withAIBrixBentoMLCTranslate2ExLlamaV2ExLlamaV3exofastllmFlashInfergpu-unlockKTransformersLitServeLMCacheLMDeployLocalAImistral.rsMLC LLMMooncakeSGLangText Generation WebUItinygradTransformersTriton Inference ServervLLMXinferenceFine-tune withAccelerateAlignment HandbookAxolotlColossal-AIDeepSpeedEasyR1FlashAttentionH2O LLM StudioJAXKerasLiger KernelLitGPTLLaMA-FactoryMegatron-LMms-swiftnanotronOLMo-coreOpenRLHFOumiPEFTPyTorch Lightningsd-scriptsSimpleTunerTorchTitantorchtuneTransformer EngineTRLUnslothverlConvert withAQLMAutoRoundCore ML ToolsExecuTorchGPTQModelLLM CompressormergekitNeural CompressorOliveONNXOptimum QuantoOptimumsafetensorsTensorRT Model OptimizertorchaoEvaluate withAutoevalsDeepEvalEvaluateHELMLightEvallm-evaluation-harnessOpenCompassOpikRagassimple-evalsSWE-benchTruLensWeaveScan withAgentDojoAI Fairness 360DeepTeamDetoxifyFairlearngarakGiskardGuardrailsMCP ScannerModelScanNeMo GuardrailspicklescanpromptfooWorks withAgent Development KitAG2Google AI Edge GalleryAiderany-llmAnythingLLMBAMLbrowser-useCAMELHugging Face Chat UIChatboxCherry StudioClineCodex CLICogneeCrewAIDeerFlowDSPyEnchantedGemini CLIGitHub MCP ServergooseGPT ResearcherGradioGraphitiGuidanceHollamaInstructorKilo CodeLangflowLangGraphLiteLLMLlamaIndexLLMLocalAGILoLLMs WebUIMCP Python SDKMem0MetaGPTNextChatOllama JavaScriptOllama PythonOpenAI Agents SDKopencodeOpenHandsotermOutlinesPage AssistPlaywright MCPPortkey AI GatewayPrivateGPTQwen CodeRWKV RunnerSemantic KernelsmolagentsStrands AgentsVaneXGrammar
OverviewModel cardFilesUse itLicenseProvenanceModel tree

Overview

Size
224 GB
Weights
224 GB (safetensors)
Parameters
572.1B
Formats
safetensors
Task
text generation
License
MIT (permissive)
Files
52, each with a sha256
Publisher
mlx-community
Added
2026-11-15
Section
alt.freeweights.text
Lineage

Original release from the publisher.

Base model: inclusionAI/Ling-3.1-flash-1019

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 05:15 UTCpassed

Model card

From the publisher at commit 3f47161. Shown as plain text with links; images open at the publisher.

Ling-3.1-flash-1019-oQ2.5e-mtp#

This model was quantized using oQ (oMLX v0.7.3rc1) mixed-precision quantization.

A mixed-precision MLX checkpoint of inclusionAI/Ling-3.1-flash-1019, an upgraded version of Ling-3.1-flash with improved reasoning and agentic capabilities. It is a hybrid-linear mixture-of-experts model with 560B total and 25B active parameters, built from 63 KDA linear-attention layers and 9 gated MLA layers, with 512 routed experts of which 8 are active per token. It includes the language model and the MTP (NextN) layer, which oMLX runs as Lightning MTP speculative decoding.

The effective average is 3.132 bits/weight for the complete checkpoint.

Quantization and Bit Distribution#

oQ2.5e allocates bits by measured layer sensitivity, then rounds each group with an importance matrix (256 samples x 512 tokens, collected layer by layer from the original FP8 weights) and a weighted least-squares refit of its scale and bias.

Storage formatLogical weightsShare of totalTensor storageEffective bits/weight
Affine 2-bit, group 64300.648B52.550%87.500 GiB2.50
Affine 3-bit, group 64246.961B43.166%100.625 GiB3.50
Affine 4-bit, group 648.053B1.408%4.219 GiB4.50
Affine 8-bit, group 1281.085B0.190%1.042 GiB8.25
Affine 8-bit, group 6415.333B2.680%15.173 GiB8.50
BF160.041B0.007%0.076 GiB16.00
FP320.000B0.000%0.000 GiB32.00
Total572.121B100%208.635 GiB3.132

Storage Breakdown#

ComponentLogical weightsStorage
Language backbone563.946B219.331 GB / 204.267 GiB
MTP head8.174B4.690 GB / 4.367 GiB
Total572.121B224.020 GB / 208.635 GiB

License#

MIT, following the source model.

Files

52 files · 224 GBWeights 224 GB (safetensors). Sizes and sha256 from the signed manifest.
Show all 52 files with their sha256
FileSizesha256
.gitattributes1.57 KB34448b82c17d60fec9b65b1f093c115ddbaadc04beb1b0140b6bfed2e012a930
README.md2.20 KBd5d2a421c29eede4a21141230be57cbbb2d1da2fe9ad18510b94a02c2c5d352c
chat_template.jinja6.04 KBf9beb6cdd406923de4999458056f2cccd016f348ccfa588c0bdb82ee0dbce457
config.json213 KB48faafa2a2294a3dfb34c71795ff782ed6bbeef71125d20e0e6ed3489fbda7d9
configuration_bailing_moe_v3.py4.72 KBf2c048966aec8a2f042cfeb1351f74d51a28589b409c55baae7d24e841c1f6c4
model-00001-of-00041.safetensors5.02 GB09d3843efddd071d014116f61cc4a4bda5dbca906d7ef80114d66d31f53125e3
model-00002-of-00041.safetensors5.03 GB639de27ad8cd1090db2cc8777caf64c282d76bcc9404b83900a37d0682afe279
model-00003-of-00041.safetensors5.00 GB3f0ff55543abd47d3576c99650e45e1132d49927acd40dded91cd9f525b9102e
model-00004-of-00041.safetensors5.05 GB920d30f9e643d55ae6eaf24552fab5225ef9cc3ecc99f5968dd123f7726e592e
model-00005-of-00041.safetensors5.71 GB6d8d8e072c9d14309bb449965664cce5987846f4b7805a5ca44704869213d1ae
model-00006-of-00041.safetensors5.72 GBfb8e74807a626503cd2a52f7d97438fe36a1391e1176ca95113c2d2b88f4a9e8
model-00007-of-00041.safetensors5.71 GBb2993bf5fd9da6ba816285b078003847c77c6a548c3a66603463753442a28944
model-00008-of-00041.safetensors5.71 GBd189c7e75de7892b5dbd575e93634819d992267aff9031eaa83d73f65ff5f096
model-00009-of-00041.safetensors5.71 GB0004829f99160a219f343f9eb8e849716293b3ce2f6ed5348751bceea0838a13
model-00010-of-00041.safetensors5.72 GB3c72dea899303dd38b55d4acf06cdb877c8bc99c59042e9fa033a7ab53c98d5f
model-00011-of-00041.safetensors5.71 GBd994f27b3ed3060a3ff0e1984d90eda432d5e769b5a1907e8527df3a47cddb68
model-00012-of-00041.safetensors5.71 GB2cc39ce57115175a26d23c1f2ce0bb08a5dec0e0ff746e500b212073c1ef7e04
model-00013-of-00041.safetensors5.71 GBfc88c1525797bf84b4f9546f647e5bdebd10316f619ea79f8b585078503df8b0
model-00014-of-00041.safetensors5.72 GB3659d9b4e74e9109161149369f159b1c77f839e4bac046dd0c84f1ef310bef26
model-00015-of-00041.safetensors5.71 GBb9d387898bf9a3ca773ead44a7a976d2d2965f2a5e6995e203f039b253bb033e
model-00016-of-00041.safetensors5.71 GB1d2e9b9a2cd3e58620681707579cc93e78ca0c73a8b151710b17df4d6ce756b7
model-00017-of-00041.safetensors5.88 GB86ca0969faa3664ea8a21d9becf573d0544264e550e4e34872f80e4642e95c9f
model-00018-of-00041.safetensors5.55 GBa078677574d096f387541d18fa36d441a32ce95299b79876af1466d0876ade99
model-00019-of-00041.safetensors5.71 GB65e2efc501e2b1e791dba93001ebf0c747361304cb934bd0a25247e05d1184db
model-00020-of-00041.safetensors6.04 GB5d676ccff94f983fab99b762201357789d6b7887facc0e110ee87ab1343977e3
model-00021-of-00041.safetensors5.20 GB612f3476aeea5918da163fef75e053dd8f70ee622abcc4a01680b8c91b419e84
model-00022-of-00041.safetensors5.21 GB2a43df18c4908e794af0c614e30ae34b30e4c182592ea6d025b2a8b09422c973
model-00023-of-00041.safetensors5.89 GBeb7445deb0326fa7a56bfc634a09088e05007d93579486e815ac4bd756832706
model-00024-of-00041.safetensors5.87 GBd9d4291bb6bc9f786fb7ab945c4ec6f66afb777251ac678135c8170efa2c4816
model-00025-of-00041.safetensors5.71 GB0637d1567e16128bad62833e0a19a8d432613dddad7baeb6182bafb2bab6cdf0
model-00026-of-00041.safetensors5.71 GB6ae3d49ce76461f12595e5f7bf59dc5a7b42892e13b2bdd851bc0bcd7e974a90
model-00027-of-00041.safetensors5.55 GB7959ea7410e696db9d8cdf89e5fc2e86ea247cf26d83e0a73e5bf57e992d4bfb
model-00028-of-00041.safetensors5.71 GB6e39fac91ae67e11b60b269fbb5731b6404e52512813e3b28dd1c69e800a8469
model-00029-of-00041.safetensors5.71 GB57bd05bea19503b99ca36c66056634fbf40090879b11c2cef94176280091f34e
model-00030-of-00041.safetensors5.71 GB6a29286c047d6c4032ba4ff0a0abffcf8612bfaace0555e8f837a826bee0a29f
model-00031-of-00041.safetensors5.72 GB4ba40bc0cd0626218de7c806add47b40b4c31a2b33a53c8253936acd6998452a
model-00032-of-00041.safetensors5.71 GBa7a8e10349854c3d578dbfbc05a371093e43535612a5d75a7928775b57a1b3dc
model-00033-of-00041.safetensors5.71 GB3f058fc74c63d7b673ff93690ee84f48c802417db92ccd216047d29f318ed131
model-00034-of-00041.safetensors5.71 GB9e6b5681bbca27f06ced93d4177fa5f30158532c4b3c677d342aadf91e8bc41f
model-00035-of-00041.safetensors5.72 GB5b7aca8bb2f299d38de41b442a5098bf6d4ec7bcb82b69ff8f2c87410c064384
model-00036-of-00041.safetensors5.71 GBcc9142b8330588401eba75ca1ee0484d0b9bec95bbadb6794b129d55747dcc3c
model-00037-of-00041.safetensors5.21 GB1a73a47cecba188ef003ab90813b9321eff36db53f1d3f41cfdb2555683c1078
model-00038-of-00041.safetensors5.54 GBec735169c856fb1143554150f798067192c69c1eecd73e6190d3cf5e79b0e638
model-00039-of-00041.safetensors5.54 GB6fd1ac4f82cf96e68cc1d7574de4621db339a5e0a7585e4bc7a902492b0bcdee
model-00040-of-00041.safetensors5.38 GB245fbedce330305ea7f61c9329a3ad9760911a92e85b7a07f66e11320ed5a091
model-00041-of-00041.safetensors11.1 MBd897137d578c94f7513d93abfc665dd7f865e8a95391434a89ab52dbeea3c32e
model.safetensors.index.json314 KB2a950e320dff5b5410e670d55b9b3ecd369b6c9d13089ff80f38c286e66ea167
modeling_bailing_moe_v3.py74.1 KB8722d34e4617db0d344b925fa2f1157434603f44424bd68107403d456ad31a74
oq_imatrix_report.json50.0 KBf07fb6e79991414d214547f39ff6d156faa5909fe6c8aae44620bee13460e961
special_tokens_map.json580 B69b63b9f81044ead642d16a5fdc01bcc737dc1183746485c8397ab14d3126614
tokenizer.json12.2 MB40fb9d7d7795b8bd305aeff39ce9963f3f450915b9553f2938e009be9a1fed60
tokenizer_config.json49.7 KB616f1e4ef8474790ea3ac6781a9de2df33f89e0c94a62cefaff7db1808326db1
VersionsEach time the pinned upstream commit or file list changed in this repo. Newest first.
Upstream commitDateContents
3f471617f3 pinned now2026-11-15 05:15 UTC52 files · 224 GB

Use it

$ fw get mlx-community/Ling-3.1-flash-1019-oQ2.5e-mtp
# manifest sha256 e68fc13f9901… · 224 GB · 52 files
$ fw verify ./Ling-3.1-flash-1019-oQ2.5e-mtp mlx-community/Ling-3.1-flash-1019-oQ2.5e-mtp
fw.pyChecks the signature and every sha256, tries each source in order, resumes.
$ python3 fw.py get https://freeweights.allternit.com/manifests/mlx-community/Ling-3.1-flash-1019-oQ2.5e-mtp.json --dir ./Ling-3.1-flash-1019-oQ2.5e-mtp $ python3 fw.py verify ./Ling-3.1-flash-1019-oQ2.5e-mtp ./mlx-community__Ling-3.1-flash-1019-oQ2.5e-mtp.json
hfThe Hugging Face CLI with one setting. Same repo id.
$ HF_ENDPOINT=https://freeweights.allternit.com hf download mlx-community/Ling-3.1-flash-1019-oQ2.5e-mtp --revision 3f471617f3031e610bf289f78d563e1aa4cc13eb --local-dir ./Ling-3.1-flash-1019-oQ2.5e-mtp
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/Ling-3.1-flash-1019-oQ2.5e-mtp", revision="3f471617f3031e610bf289f78d563e1aa4cc13eb") model = AutoModelForCausalLM.from_pretrained("mlx-community/Ling-3.1-flash-1019-oQ2.5e-mtp", revision="3f471617f3031e610bf289f78d563e1aa4cc13eb", device_map="auto")
vLLMServe an OpenAI-compatible API.
$ HF_ENDPOINT=https://freeweights.allternit.com vllm serve mlx-community/Ling-3.1-flash-1019-oQ2.5e-mtp --revision 3f471617f3031e610bf289f78d563e1aa4cc13eb
Fine-tune in Studio

License

Name
MIT License
Id
MIT
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
d5d2a421c29eede4a21141230be57cbbb2d1da2fe9ad18510b94a02c2c5d352c
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: Ling-3.1-flash-1019-oQ2.5e-mtpDate: 2026-11-15Commit: 3f47161Signed: 2DC28CC7B26A283B ✓
Upstream
mlx-community/Ling-3.1-flash-1019-oQ2.5e-mtp
Pinned commit
3f471617f3031e610bf289f78d563e1aa4cc13eb
Manifest
Ling-3.1-flash-1019-oQ2.5e-mtp.json · .minisig
Manifest sha256
e68fc13f99017f5886e01846eb49e46032816c9f0bb1b42abb820204705834c8
Signature
minisign, Ed25519, key 2DC28CC7B26A283B
Created
2026-11-15 05:15 UTC
Generator
freeweights-manifest/0.1.0
Takedowns
None.

How to verify

$ curl -O https://freeweights.allternit.com/manifests/mlx-community/Ling-3.1-flash-1019-oQ2.5e-mtp.json $ curl -O https://freeweights.allternit.com/manifests/mlx-community/Ling-3.1-flash-1019-oQ2.5e-mtp.json.minisig $ minisign -Vm Ling-3.1-flash-1019-oQ2.5e-mtp.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

inclusionAI/Ling-3.1-flash-1019 · this model is a quantization of it

No model in the catalog names this one as its parent.

Get it
Download
$ HF_ENDPOINT=https://freeweights.allternit.com hf download mlx-community/Ling-3.1-flash-1019-oQ2.5e-mtp
Open in Freeweights Studio

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

Signed manifest
minisign, key 2DC28CC7B26A283B. Ling-3.1-flash-1019-oQ2.5e-mtp.json. Provenance. Takedown history: none.
Source
huggingface.co/mlx-community/Ling-3.1-flash-1019-oQ2.5e-mtp

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