Skip to content
AI Atlas
ArtifactquantizationBF16/BOOL/I64Open weightsof Qwen/Qwen3-Coder-Next-FP8

togethercomputer/Aurora-Spec-Qwen3-Coder-Next-FP8

published by Together AIhuggingface.co/togethercomputer/Aurora-Spec-Qwen3-Coder-

This is a quantization of Qwen/Qwen3-Coder-Next-FP8, not an independent model. Parameters, benchmarks, prices and lineage are recorded on the canonical model. Open Qwen/Qwen3-Coder-Next-FP8 →

Updated 1 h ago · first seen 12 Sept 2026

artifact_01M2AS5MM50K2P9ARBM4E4V719

File size
Format
Downloads
Published

Artifact facts

Weights dtype

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium

File size

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium

Downloads

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium

Likes

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium

Hugging Face repo

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium

Base model

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium

Pipeline tag

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium

Access

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium

License

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium

Release date

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium

Last modified

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium

Tags

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium

Hardware fit (this packaging)

Estimated

37 of 37 device × quantization combinations fit.

Run locally: your machine →
Estimated hardware fit
HardwareQuantizationDevice memoryEst. memoryFits
Apple M3 Ultra4bit—0.8 GB est.Yes
Mac Studio (Apple M5 Ultra)4bit—0.8 GB est.Yes
Apple M2 Ultra4bit—0.8 GB est.Yes
Apple M1 Ultra4bit—0.8 GB est.Yes
Apple M3 Max4bit—0.8 GB est.Yes
Apple M4 Max4bit—0.8 GB est.Yes
Mac Studio (Apple M5 Max)4bit—0.8 GB est.Yes
MacBook Pro (Apple M5 Max)4bit—0.8 GB est.Yes
Apple M2 Max4bit—0.8 GB est.Yes
Apple M1 Max4bit—0.8 GB est.Yes
Apple M4 Pro4bit—0.8 GB est.Yes
Mac mini (Apple M5 Pro)4bit—0.8 GB est.Yes
MacBook Pro (Apple M5 Pro)4bit—0.8 GB est.Yes
Apple M3 Pro4bit—0.8 GB est.Yes
Apple M1 Pro4bit—0.8 GB est.Yes
Apple M2 Pro4bit—0.8 GB est.Yes
Apple M44bit—0.8 GB est.Yes
iMac (Apple M4)4bit—0.8 GB est.Yes
Mac mini (Apple M6)4bit—0.8 GB est.Yes
MacBook Air (Apple M5)4bit—0.8 GB est.Yes
MacBook Pro (Apple M5)4bit—0.8 GB est.Yes
Apple M24bit—0.8 GB est.Yes
Apple M34bit—0.8 GB est.Yes
Apple M14bit—0.8 GB est.Yes
NVIDIA GeForce RTX 30904bit24 GB0.8 GB est.Yes
NVIDIA GeForce RTX 40904bit24 GB0.8 GB est.Yes
NVIDIA GeForce RTX 50904bit32 GB0.8 GB est.Yes
NVIDIA A100 80GB4bit80 GB0.8 GB est.Yes
NVIDIA H100 SXM4bit80 GB0.8 GB est.Yes
NVIDIA H100 NVL4bit94 GB0.8 GB est.Yes
NVIDIA DGX Spark4bit128 GB0.8 GB est.Yes
NVIDIA H2004bit141 GB0.8 GB est.Yes
NVIDIA H200 NVL4bit141 GB0.8 GB est.Yes
NVIDIA B2004bit180 GB0.8 GB est.Yes
AMD Instinct MI300X4bit192 GB0.8 GB est.Yes
AMD Instinct MI325X4bit256 GB0.8 GB est.Yes
NVIDIA DGX B2004bit1,440 GB0.8 GB est.Yes
Assumptions (7)
  • Estimated, not measured: weights = parameters × bytes/param × 1.15 runtime overhead (or the observed artifact file size when one is recorded).
  • bytes/param: 4bit = 0.5, 8bit = 1.0, fp16 = 2.0 (uniform quantization, no per-layer exceptions).
  • KV cache: 2 × layers × kv_heads × head_dim × 2 bytes × context × batch when the architecture is known; otherwise 0.5 GB per 8 192 tokens (× batch), independent of architecture (GQA/MLA models need less).
  • A model 'fits' when the estimate is at most the device memory minus 2 GB reserved for the OS and framework.
  • Mixture-of-experts models are estimated on total parameters (all experts must be resident); active parameters are ignored.
  • Device memory uses the largest configuration when several are listed (e.g. Apple silicon tiers).
  • Multi-GPU: device memories are summed; interconnect bandwidth, tensor-parallel replication and pipeline bubbles are not modelled.

Timeline

Full timeline →

Provenance

Attributed facts

27

Source tiers

T227

Freshest observation

2 h ago

Conflicts

None

Source documents
SourceDocumentTypeTierLast observedSnapshots
Hugging Face Hub (public pages, model cards, papers)huggingface.co/togethercomputer/Aurora-Spec-Qwen3-Coder-Next-FP8 model_pageT2· Quality secondary2 h ago1
Hugging Face Hub (public pages, model cards, papers)huggingface.co/models?author=togethercomputer&p=0&sort=downloads listingT2· Quality secondary2 h ago8

Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.