Skip to page content
SAVRN
Search Plan a project

Open-weight model · Text generation

Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-MLX

by Elvis McFee Ecmcfee/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-MLX

Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-MLX is an open-weight model for text generation from Elvis McFee, released under Apache License 2.0. It has 27.3B parameters and a 262,144-token context. At 16-bit it needs about 65.4 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index.

MLX-optimized version of DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP for Apple Silicon inference. Converted from the Q80 GGUF quantization using gguf2mlx.

Parameters27.3B
Context262,144
Weights54.6 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads—

Runs On

What it takes to serve Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-MLX (27.3B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

PrecisionWeightsMemory neededCheapest setupPer hourAlso fits
16-bit 54.5 GB 65.4 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
8-bit 27.3 GB 32.7 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
4-bit 13.6 GB 16.4 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00

Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Oct 16, 2026.

Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-MLX on every accelerator the SAVRN Index prices, at every precision

Model Card

By Elvis McFee, published under apache-2.0, revision 27dfa6fcde6f.

MLX-optimized version of DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP for Apple Silicon inference. Converted from the Q80 GGUF quantization using gguf2mlx. All 866 tensors are dequantized to float16, so this model has identical quality to the original GGUF weights — no precision loss from the conversion. Apache-2.0 (same as the original model)

Read Elvis McFee's full model card

Qwen3.8-27B-TurboFCFusion-735-882-Here-Uncen-NEO-CODER-MAX-MTP (MLX)

MLX-optimized version of DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP for Apple Silicon inference.

Converted from the Q8_0 GGUF quantization using gguf2mlx. All 866 tensors are dequantized to float16, so this model has identical quality to the original GGUF weights — no precision loss from the conversion.

Inference

Run with MLX-LM:

python -m mlx_lm.chat --model <your-username>/Qwen3.8-27B-TurboFCFusion-735-882-Here-Uncen-NEO-CODER-MAX-MTP-MLX

Or from Python:

from mlx_lm import load, generate

model, tokenizer = load("<your-username>/Qwen3.8-27B-TurboFCFusion-735-882-Here-Uncen-NEO-CODER-MAX-MTP-MLX")
prompt = "Write a short story about a robot learning to dream."
response = generate(model, tokenizer, prompt, max_tokens=512)
print(response)

Conversion Notes

  • Source: Qwen3.8-27B-TurboFCFusion-735-882-Here-Uncen-NEO-CODER-MAX-MTP-Q8_0.gguf (30.24 GB)
  • Target: safetensors (float16, ~55 GB total across 12 shards)
  • Architecture: Qwen3.5 dense (65 layers, 5120 hidden, 24 heads, 248320 vocab)
  • Tokenizer: GPT-2 style (248320 tokens)
  • Chat template: Qwen3 standard

License

Apache-2.0 (same as the original model)

Configuration

Architecture
Qwen3ForCausalLM
Context length (tokens)
262,144
Layers
64
Hidden size
5,120
Feed-forward size
17,408
Attention heads
24
Key/value heads
4
Vocabulary size
248,320
RoPE base
1e+07
Stored precision
float16
Model type
qwen3

Identity and Version

Repository
Ecmcfee/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-MLX
Publisher
Elvis McFee
Task
Text generation
Modality
Text
Library
mlx
Parameters
27.3B parameters
Languages
en, zh
Revision
27dfa6fcde6f415cdae68815c5b161c788c35042
First published
2026-10-11
Last updated
2026-10-11

Files and Weights

23 files, 54.7 GB in total. The weights are 13 files totalling 54.6 GB in safetensors.

Weights13 files · 54.6 GB
Configuration3 files · 63.2 KB
Tokenizer4 files · 22.4 MB
Documentation1 file · 2.0 KB
Other1 file · 9.0 KB
Repository1 file · 1.6 KB
Every file
FileTypeSizeSHA-256
model-00001-of-00012.safetensorsWeights5.1 GB f55a6994a042
model-00002-of-00012.safetensorsWeights4.5 GB 80f809976951
model-00003-of-00012.safetensorsWeights4.6 GB 73551853ca77
model-00004-of-00012.safetensorsWeights4.5 GB 2e5f0952a596
model-00005-of-00012.safetensorsWeights4.6 GB 6f9b46bda600
model-00006-of-00012.safetensorsWeights4.5 GB 77e1dc9ec5ec
model-00007-of-00012.safetensorsWeights4.6 GB 78da1a43400d
model-00008-of-00012.safetensorsWeights4.5 GB 0ec753f971e2
model-00009-of-00012.safetensorsWeights4.6 GB e74b6f59d33b
model-00010-of-00012.safetensorsWeights4.5 GB f6c3cec26030
model-00011-of-00012.safetensorsWeights4.6 GB 39f771fa290c
model-00012-of-00012.safetensorsWeights3.8 GB ed07c2cf1358
mtp.safetensorsWeights104.9 MB 14eb6a9b3892
config.jsonConfiguration695 B —
model.safetensors.index.jsonConfiguration62.4 KB —
special_tokens_map.jsonConfiguration95 B —
README.mdDocumentation2.0 KB —
chat_template.jinjaOther9.0 KB —
.gitattributesRepository1.6 KB —
merges.txtTokenizer3.4 MB —
tokenizer.jsonTokenizer12.8 MB 0516fcfed2ba
tokenizer_config.jsonTokenizer9.5 KB —
vocab.jsonTokenizer6.2 MB —

License and Download

License
apache-2.0
Access
Open weights, no gate
Download size
54.6 GB
Download from Elvis McFee

Released by Elvis McFee through its official repository on Hugging Face. Read the license.

Built From

Memory Requirements

PrecisionWeights in memory
As published54.6 GB
16-bit54.5 GB
8-bit27.3 GB
4-bit13.6 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Questions About Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-MLX

How much GPU memory does Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-MLX need?

About 65.4 GB at 16-bit and 16.4 GB at 4-bit: the weights (27.3B parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-MLX on?

At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.

Can I use Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-MLX commercially?

Yes. Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-MLX is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.

What is Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-MLX's context length?

262,144 tokens, from the maximum position embeddings in its published configuration.

Similar Models

Benefits high quality CPU inference TQ2 on Llama.cpp and Ollama via QAT - Robotcs, Routing, Coding, Multimedia, Advanced tool calling via JiRackDeltaNetTokenizer - JiRack DeltaNet understand video and images that best for Robotics also A fast and efficient 27B model optimized for CPU inference. Built on a Qwen3.8-style DeltaNet architecture (hybrid attention + SSM), with an updated tokenizer that includes Routing, Media, Vision, Sound, Tool call, and Robotics tags. Ready-to-run GGUF quantizations, and native Ollama support with reasoning disabled by default for fast, direct responses. - JiRack is a cloud-ready model that helps save money on cloud infrastructure. It can be used as an expert…

Open weights mit 27.3B parameters 262,144 tokens
View model

Model · Text generation

Ternary-Bonsai-27B-mlx-2bit

Prism ML

Full 27B-class reasoning in ternary transformer weights — on everyday laptops - \~7.2 GB deployed footprint (down from \~54 GB FP16) — full 27B-class reasoning on a standard laptop or a single GPU - 95% of FP16 intelligence retained: 80.49 average across 15 thinking-mode benchmarks — a higher score than the conventional IQ2XXS build (72.73) at less than two-thirds of its footprint - Retains thinking, reasoning, and agentic behavior deep in the sub-4-bit regime, where conventional low-bit representations collapse: math within two points of full precision (93.40), coding at 85.96, agentic tool use at 74.01 - End-to-end ternary language weights across embeddings, attention projections, MLP…

Open weights apache-2.0 27.4B parameters 262,144 tokens mlx
View model

Qwen3.8-cyber-RedTeam-Surgical-Abliterated (27B) is an unconstrained foundation engine engineered for autonomous cybersecurity agents, vulnerability research, and offensive cyber operations. Built for executing low-level technical directives rather than conversational chatting, it features complete refusal orthogonalization, native multi-step tool calling, and high-throughput linear-attention efficiency: - Quantization Rationale (Native FP8 vs. 4-bit INT4/AWQ): Evaluated across adversary emulation and systems engineering tasks: A standalone orchestration utility is included for automated environment bootstrapping, weight synchronization, and hardware-adaptive server initialization: This…

Open weights apache-2.0 27.4B parameters 262,144 tokens transformers
View model

Prism ML's ternary Ternary-Bonsai-2-27B build of Qwen/Qwen3.8-27B, repacked for chad, a Claude-Code-style local coding agent for Apple Silicon, with its speculative decoder bundled in. This is chad's default model. Created using Bonsai by Prism ML. with, already quantized. Nothing is built on first run. Every projection of Qwen3.8-27B (a dense qwen35 hybrid: 64 layers, 48 GatedDeltaNet + 16 full attention) is stored in a Hadamard-rotated basis: multiplied by a fixed sign vector and put through a blockwise Walsh-Hadamard transform offline, then quantized to 2-bit affine group-128 whose three levels reproduce the ternary set {−s, 0, +s}. The rotation costs no extra bits and no extra weight…

Open weights apache-2.0 26.9B parameters 262,144 tokens mlx
View model

Model · Text generation

Darwin-27B-RSI

FINAL_Bench

Darwin-27B-RSI is Darwin-27B-Opus after Recursive Self-Improvement (RSI): the model was improved using only signal it produced itself. During self-improvement, the model itself (its weights) improves by learning only from its own solutions. No human-written solutions or reasoning traces are used; correctness is checked automatically (agreement across its own samples and code execution). Under an identical evaluation protocol, Darwin-27B-RSI improves over its parent on graduate-level science reasoning — +5.24 points on GPQA Diamond (single sample) and +3.79 points with majority voting — with every gain statistically significant in paired tests. As the reasoning engine of Darwin-27B-JEV on…

Open weights apache-2.0 26.9B parameters 262,144 tokens transformers
View model

Model · Text generation

Qwen3.8-27B-pi-mlx-6Bit

SolsticeAI

The Model SolsticeAI/Qwen3.8-27B-pi-mlx-6Bit was converted to MLX format from bytkim/Qwen3.8-27B-pi using mlx-lm version 0.31.2.

Open weights apache-2.0 26.9B parameters 262,144 tokens transformers
View model