Skip to content
AI Atlas
ModelDeprecated

Qwen3 8B

Qwenhuggingface.co/Qwen/Qwen3-8B

Qwen3-8B is a dense 8.2B parameter causal language model from the Qwen3 series, designed for both reasoning-heavy tasks and efficient dialogue. It supports seamless switching between "thinking" mode for math,...

Open in Graph
data quality68

Updated 4 h ago · first seen 11 Sept 2026

model_01M294WWGKJRNVC964ZQYXDZR1

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 6 h agomedium

Status

Source:Artificial AnalysisT2observed 9 h agomedium

Knowledge cutoff

Source:OpenRouter public model & pricing listingT2observed 11 h agomedium

Model card
huggingface.co/Qwen/Qwen3-8B

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 11 h agomedium

Openrouter id

Source:OpenRouter public model & pricing listingT2observed 11 h agomedium

Architecture

Architecture

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

Model type

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

Parameters

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 11 h agomedium

Tokenizer

Source:OpenRouter public model & pricing listingT2observed 11 h agomedium

Weights dtype

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

File size

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

Library name

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

Pipeline tag

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 11 h agomedium

Hugging Face repo

Source:OpenRouter public model & pricing listingT2observed 11 h agomedium

Capabilities

Modalities

Modalities
text
Input
text
Output
text

Capabilities

  • Tool calling

    Yes

    OpenRouter public model & pricing listing · T2

  • Structured output

    Yes

    OpenRouter public model & pricing listing · T2

  • Reasoning

    Yes

    OpenRouter public model & pricing listing · T2

  • Vision

    Unavailable

  • Audio

    Unavailable

  • Fine-tuning available

    Unavailable

Context window

Source:OpenRouter public model & pricing listingT2observed 11 h agomedium

Max output

Source:OpenRouter public model & pricing listingT2observed 11 h agomedium

Knowledge cutoff

Source:OpenRouter public model & pricing listingT2observed 11 h agomedium

Tokenizer

Source:OpenRouter public model & pricing listingT2observed 11 h agomedium

Benchmarks9

Compare with another model →

Comparable same task and conditions · Partially comparable same task, conditions differ (effort, temperature, judge) · Not comparable different variant or metric

No benchmark results recorded

Results appear when a tier 1–3 source publishes them; we never copy scores without a source.

Providers & Pricing1

All offers in the price terminal →
Current prices per 1M tokens
ProviderInput / 1MOutput / 1MCached inBatch in / outContextObservedSource
OpenRouterqwen/qwen3-8b$0.117$0.455——131.1K3 h agoopenrouter.aiT2

USD per 1M tokens as published by each provider (USD). Rows are append-only: every change is kept in the history below.

Price history

Step lines per provider; amber markers are recorded changes. Click a marker or a row for the evidence behind that price.

Output price · USD / 1M tokens 1 provider

Output price history of Qwen3 8B$0$0.20$0.40$0.60Sept 26Sept 26Sept 26Sept 26Sept 26Sept 26Sept 26Sept 26OpenRouter: first observed → $0.455 · 11 Sept 2026
  • OpenRouter
  • OpenRouterfirst observed $0.45511 Sept 2026

Input price · USD / 1M tokens 1 provider

Input price history of Qwen3 8B$0$0.05$0.10$0.15Sept 26Sept 26Sept 26Sept 26Sept 26Sept 26Sept 26Sept 26OpenRouter: first observed → $0.117 · 11 Sept 2026
  • OpenRouter
  • OpenRouterfirst observed $0.11711 Sept 2026

Hardware fit37

Estimated

37 of 37 device × quantization combinations fit.

Run locally: your machine →
Estimated hardware fit
HardwareQuantizationDevice memoryEst. memoryFits
Apple M3 Ultra4bit—5.2 GB est.Yes
Mac Studio (Apple M5 Ultra)4bit—5.2 GB est.Yes
Apple M2 Ultra4bit—5.2 GB est.Yes
Apple M1 Ultra4bit—5.2 GB est.Yes
Apple M3 Max4bit—5.2 GB est.Yes
Apple M4 Max4bit—5.2 GB est.Yes
Mac Studio (Apple M5 Max)4bit—5.2 GB est.Yes
MacBook Pro (Apple M5 Max)4bit—5.2 GB est.Yes
Apple M2 Max4bit—5.2 GB est.Yes
Apple M1 Max4bit—5.2 GB est.Yes
Apple M4 Pro4bit—5.2 GB est.Yes
Mac mini (Apple M5 Pro)4bit—5.2 GB est.Yes
MacBook Pro (Apple M5 Pro)4bit—5.2 GB est.Yes
Apple M3 Pro4bit—5.2 GB est.Yes
Apple M1 Pro4bit—5.2 GB est.Yes
Apple M2 Pro4bit—5.2 GB est.Yes
Apple M44bit—5.2 GB est.Yes
iMac (Apple M4)4bit—5.2 GB est.Yes
Mac mini (Apple M6)4bit—5.2 GB est.Yes
MacBook Air (Apple M5)4bit—5.2 GB est.Yes
MacBook Pro (Apple M5)4bit—5.2 GB est.Yes
Apple M24bit—5.2 GB est.Yes
Apple M34bit—5.2 GB est.Yes
Apple M14bit—5.2 GB est.Yes
NVIDIA GeForce RTX 30904bit24 GB5.2 GB est.Yes
NVIDIA GeForce RTX 40904bit24 GB5.2 GB est.Yes
NVIDIA GeForce RTX 50904bit32 GB5.2 GB est.Yes
NVIDIA A100 80GB4bit80 GB5.2 GB est.Yes
NVIDIA H100 SXM4bit80 GB5.2 GB est.Yes
NVIDIA H100 NVL4bit94 GB5.2 GB est.Yes
NVIDIA DGX Spark4bit128 GB5.2 GB est.Yes
NVIDIA H2004bit141 GB5.2 GB est.Yes
NVIDIA H200 NVL4bit141 GB5.2 GB est.Yes
NVIDIA B2004bit180 GB5.2 GB est.Yes
AMD Instinct MI300X4bit192 GB5.2 GB est.Yes
AMD Instinct MI325X4bit256 GB5.2 GB est.Yes
NVIDIA DGX B2004bit1,440 GB5.2 GB est.Yes
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.
ANCESTORS 1DESCENDANTS 3 · ARTIFACTSQwen3-8B-BaseQwenQwen3-8B-Base — QwenGDN-primed-HQwen3-8B-Instruct8.5BGDN-primed-HQwen3-8B-Instruct — 8.5BGKA-primed-HQwen3-8B-Reasoner8.5BGKA-primed-HQwen3-8B-Reasoner — 8.5BMolmo2-8B8.66BMolmo2-8B — 8.66B1 quantizationartifacts · collapsed1 quantization — artifacts · collapsedQwen3 8B8.19B params · this modelQwen3 8B — 8.19B params · this model

Papers2

Timeline13

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

Base modelbase_model1

Claim history for Base model
ValueValid from → toStatusSourceConfidenceExtractor
Qwen/Qwen3-8B-Base → 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

38

Source tiers

T238

Freshest observation

4 h ago

Conflicts

None

Source documents 10

Source documents
SourceDocumentTypeTierLast observedSnapshots
Qwen — official blogqwenlm.github.io/blog/qwen3 newsT1· Official11 h ago1
OpenRouter public model & pricing listingopenrouter.ai/api/v1/models catalogueT2· Quality secondary3 h ago6
Hugging Face Hub (public pages, model cards, papers)huggingface.co/Qwen/Qwen3-8B/raw/main/README.md model_cardT2· Quality secondary4 h ago1
Hugging Face Hub (public pages, model cards, papers)huggingface.co/Qwen/Qwen3-8B model_pageT2· Quality secondary5 h ago5
Hugging Face Hub (public pages, model cards, papers)huggingface.co/mlx-community/Qwen3-8B-4bit model_pageT2· Quality secondary5 h ago5
Hugging Face Hub (public pages, model cards, papers)huggingface.co/amazon/GDN-primed-HQwen3-8B-Instruct model_pageT2· Quality secondary5 h ago6
Hugging Face Hub (public pages, model cards, papers)huggingface.co/amazon/GKA-primed-HQwen3-8B-Reasoner model_pageT2· Quality secondary5 h ago6
Hugging Face Hub (public pages, model cards, papers)huggingface.co/allenai/Molmo2-8B model_pageT2· Quality secondary5 h ago6
Hugging Face Hub (public pages, model cards, papers)huggingface.co/models?author=Qwen&p=0&sort=downloads listingT2· Quality secondary5 h ago6
Artificial Analysisartificialanalysis.ai/leaderboards/models leaderboardT2· Quality secondary9 h ago1

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

Data quality (68/100) measures how well AI Atlas knows this entity — completeness, primary-source ratio, freshness, conflicts — never how good the model is. Methodology →