Documentation Navigation
⌘K

Models Catalog

Qwen3.8-27B-NVFP4unsloth ModelsQwen GenerationHuihui-Qwen3.8-27B-abliteratedQwen3.8-27B-Heretic-Abliterated-Uncensored-GGUFQwen3.8-27B-OBLITERATEDHuihui-Qwen3.8-27B-abliterated-GGUFQwen3.8-27B-Ridge-GGUFQwen3.8-27B-Uncensored-HauhauCS-Aggressive-MTP-GGUFUpgrade paths & migrationsGlobal leaderboard

Qwen3.8-27B-NVFP4

Community Recordactive
API Model ID:
qwen38-27b-nvfp4
Developerunsloth
Parameters27B
Context Window128K tokens
LicenseApache-2.0

Qwen3.8-27B-NVFP4 is a 27B Qwen3.8 distribution packaged in NVFP4 for efficient local inference. The important product distinction is memory efficiency and deployment, not a new foundation-model benchmark profile.

Executive Summary

A 27B NVFP4 Qwen3.8 distribution for efficient local inference.

Architecture & System Overview

Its value is deployment efficiency: the same model family is packaged for lower-memory inference. Keep artifact-specific benchmark measurements separate from the base model.

Capabilities & Highlights

Key Features

27B scale
NVFP4 quantization
128K context
Local inference oriented
Architecture & Lineage

Model lineage & specification

unslothFamily: Qwen

Technical architectural specifications, context capacities, parameter distributions, and historical lineage relationships tracked for Qwen3.8-27B-NVFP4.

Model Ancestry & LineageRelease: 2026-08-15

Current ModelQwen3.8-27B-NVFP4
Model FamilyQwen
Product Tierefficient
Previous VersionNone (initial generation)
Base / Parent Modelqwen38-27b-gguf
Lifecycle & Access
open weights quantization•active

Hardware & Execution Specschat reasoning

Parameter Count27B
Active Parameters (MoE)Dense / All active
Context Window128K tokens
Supported Modalities
text
Deployment Formats
self-hostable
LicenseApache-2.0
API Endpoint IDqwen38-27b-nvfp4
Canonical Aliases
Qwen3.8-27B-NVFP4qwen38-27b-nvfp4qwen38 27b nvfp43.8-27B-NVFP4unsloth Qwen3.8-27B-NVFP4
Developer Quickstart

Getting Started

API Reference & SDKs

Select an implementation language for Qwen3.8-27B-NVFP4:

qwen3-8-27b-nvfp4-quickstart.shstatus
repository instructions required
Architecture & Tier Comparison

Comparable Models

Open Full Interactive Comparison

Side-by-side architectural and specification comparison across equivalent models in the chat reasoning category.

Qwen3.8-27B-NVFP4This Model
Parameters27B
Context Window128K tokens
Typeopen weights quantization
Taskchat reasoning
Solar Pro 4Upstage
Parameters250B
Context Window524K
Typeapi only
Taskchat reasoning
Qwen3.8-2.4T-A95BAlibaba Cloud / Qwen
Parameters2.4T total
Context Window1M tokens
Typeopen weight
Taskchat reasoning
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4nvidia
Parameters17.8B
Context Window1M tokens
Typeopen weights
Taskchat reasoning
Evaluation Suite

Verified Benchmarks

No verified benchmark scores published or recorded yet.

Modelverse indexes standardized evaluations from official research papers, vendor system cards, and independent academic leaderboards.

Commercial Rates & Token Costs

API Pricing

Last verified: Aug 22, 2026

Pricing not publicly listed.

No public pay-per-token API rates have been officially published or indexed for this model.

Expert Editorial Assessment

Modelverse Editorial Analysis

Benchmark this exact NVFP4 artifact before publishing performance numbers.

Share this model

Found this insightful? Share it with your community on Reddit, X, or copy the link.

Verification & Provenance

Sources & Provenance

1 Primary Citation

Direct references, official documentation endpoints, evaluation papers, and launch announcements referenced for Qwen3.8-27B-NVFP4.

Hugging Face Model Card

huggingface.co

https://huggingface.co/unsloth/Qwen3.8-27B-NVFP4