Skip to content
AI Atlas
ModelActive

Seed-OSS-36B-Base

ByteDancehuggingface.co/ByteDance-Seed/Seed-OSS-36B-Base

Open in Graph
quality53

Updated 4 h ago · first seen 11 Sept 2026

model_01M294XRVGHW05AX6NTSFPK9DT

Overview

Identity

Identity block not returned by the API for this entity.

Openness

Openness not classified yet — no sourced evidence to place this model in the ontology.

Key facts

Release date

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 9 h agomedium

Architecture

Architecture

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 9 h agomedium

Model type

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 9 h agomedium

Parameters

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 10 h agomedium

Weights dtype

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 9 h agomedium

File size

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 9 h agomedium

Library name

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 9 h agomedium

Pipeline tag

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 10 h agomedium

Hugging Face repo

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 10 h agomedium

Hardware fit37

Estimated

36 of 37 device × quantization combinations fit.

Run locally: your machine →
Estimated hardware fit
HardwareQuantizationDevice memoryEst. memoryFits
Apple M3 Ultra4bit—21.3 GB est.Yes
Mac Studio (Apple M5 Ultra)4bit—21.3 GB est.Yes
Apple M2 Ultra4bit—21.3 GB est.Yes
Apple M1 Ultra4bit—21.3 GB est.Yes
Apple M3 Max4bit—21.3 GB est.Yes
Apple M4 Max4bit—21.3 GB est.Yes
Mac Studio (Apple M5 Max)4bit—21.3 GB est.Yes
MacBook Pro (Apple M5 Max)4bit—21.3 GB est.Yes
Apple M2 Max4bit—21.3 GB est.Yes
Apple M1 Max4bit—21.3 GB est.Yes
Apple M4 Pro4bit—21.3 GB est.Yes
Mac mini (Apple M5 Pro)4bit—21.3 GB est.Yes
MacBook Pro (Apple M5 Pro)4bit—21.3 GB est.Yes
Apple M3 Pro4bit—21.3 GB est.Yes
Apple M1 Pro4bit—21.3 GB est.Yes
Apple M2 Pro4bit—21.3 GB est.Yes
Apple M44bit—21.3 GB est.Yes
iMac (Apple M4)4bit—21.3 GB est.Yes
Mac mini (Apple M6)4bit—21.3 GB est.Yes
MacBook Air (Apple M5)4bit—21.3 GB est.Yes
MacBook Pro (Apple M5)4bit—21.3 GB est.Yes
Apple M24bit—21.3 GB est.Yes
Apple M34bit—21.3 GB est.Yes
NVIDIA GeForce RTX 30904bit24 GB21.3 GB est.Yes
NVIDIA GeForce RTX 40904bit24 GB21.3 GB est.Yes
NVIDIA GeForce RTX 50904bit32 GB21.3 GB est.Yes
NVIDIA A100 80GB4bit80 GB21.3 GB est.Yes
NVIDIA H100 SXM4bit80 GB21.3 GB est.Yes
NVIDIA H100 NVL4bit94 GB21.3 GB est.Yes
NVIDIA DGX Spark4bit128 GB21.3 GB est.Yes
NVIDIA H2004bit141 GB21.3 GB est.Yes
NVIDIA H200 NVL4bit141 GB21.3 GB est.Yes
NVIDIA B2004bit180 GB21.3 GB est.Yes
AMD Instinct MI300X4bit192 GB21.3 GB est.Yes
AMD Instinct MI325X4bit256 GB21.3 GB est.Yes
NVIDIA DGX B2004bit1,440 GB21.3 GB est.Yes
Apple M14bit—21.3 GB est.No
Assumptions (6)
  • Estimated, not measured: weights = parameters × bytes/param × 1.15 runtime overhead.
  • bytes/param: 4bit = 0.5, 8bit = 1.0, fp16 = 2.0 (uniform quantization, no per-layer exceptions).
  • KV cache approximated at 0.5 GB per 8 192 tokens of context, 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).

Lineage

Open in Graph →
Explicit derived_from / fine_tuned_from / distilled_from relations stated by sources; artifacts collapsed by kind.
DESCENDANTS 0 · ARTIFACTS1 quantizationartifacts · collapsed1 quantization — artifacts · collapsedSeed-OSS-36B-Base36.1B params · this modelSeed-OSS-36B-Base — 36.1B params · this model

    Timeline1

    Full timeline →

    Change history

    Temporal, append-only claims: a new observation closes the previous claim instead of overwriting it. Rewind the record with the as-of picker.
    1 claims · 1 propertiesShow all properties

    Last modifiedlast_modified1

    Claim history for Last modified
    ValueValid from → toStatusSourceConfidenceExtractor
    2025-08-26T02:33:44+00:00 → currentcurrentHugging Face Hub (public pages, model cards, papers)T2mediumdeterministic

    Claims are temporal and append-only: a new observation closes the previous claim (valid_to) instead of overwriting it. Conflicting claims from different sources are kept side by side and flagged — never averaged. Methodology →

    Provenance

    Attributed facts

    18

    Source tiers

    T218

    Freshest observation

    4 h ago

    Conflicts

    None

    Source documents 4

    Source documents
    SourceDocumentTypeTierLast observedSnapshots
    Hugging Face Hub (public pages, model cards, papers)huggingface.co/ByteDance-Seed/Seed-OSS-36B-Base/raw/main/README.md model_cardT2· Quality secondary3 h ago1
    Hugging Face Hub (public pages, model cards, papers)huggingface.co/NousResearch/Hermes-4.3-36B-GGUF model_pageT2· Quality secondary3 h ago5
    Hugging Face Hub (public pages, model cards, papers)huggingface.co/ByteDance-Seed/Seed-OSS-36B-Base model_pageT2· Quality secondary4 h ago4
    Hugging Face Hub (public pages, model cards, papers)huggingface.co/models?author=ByteDance-Seed&p=0&sort=downloads listingT2· Quality secondary4 h ago6

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

    Data quality (53/100) measures how well AI Atlas knows this entity — completeness, primary-source ratio, freshness, conflicts — never how good the model is. Methodology →