reasoning
Qwen3.8-2.4T-A95B
Open-weight flagship checkpoint of Qwen3.8; a text-only model that requires thinking mode, under the custom Qwen3.8-Max License.

Levelmodel
Classreasoning
Parameters2.4T
Context262,144
ReleasedAug 2026
Opennessopen weights restricted
Runs onmulti-GPU node
Statusactive
Benchmark Results
vendor-reportedindependent
Artificial Analysis Intelligence Index v4.3.2composite index
independent
40
03060
Score in index. Higher is better.
AA model page
Qwen3.8 2.4T A95Bindependent
40
03060
Score in index. Higher is better.
Artificial Analysis model page
GPQA Diamondknowledge and reasoning
vendor-reported
92.6
050100
Score in %. Higher is better.
Humanity's Last Exam (HLE)knowledge and reasoning
vendor-reported
43.6
050100
Score in %. Higher is better.
SWE-bench Procoding
vendor-reported
67.7
050100
Score in %. Higher is better.
Terminal-Bench 2.1agents and tool use
vendor-reported
86.6
050100
Score in %. Higher is better.
Where this comes from
Every score names who ran it: vendor-reported, maintainer-run or independent. A rank is shown with the population it was ranked in (gap rule 3).
Lineage
successorproduct of
Qwen3.8 Max is a child (product of)high confidencemodel cardWhere this comes from
Parent and child relations from model cards, Hugging Face base_model fields and the genealogy tables of the research notes. The tree shows two generations each way, 3 models. Confidence and evidence are the notes' own.
R&D Topics
Identity
- Developer
- Alibaba (Qwen)
- Family
- Qwen3.8
- Released
- Aug 2026 sourcessecondary report
- Weights published
- 8 Aug 2026
- Record checked
- 10 Oct 2026
Where this comes from
Developer and dates from the research notes. Release date is the vendor's public announcement; weights date is when open weights were published.
Architecture and Size
- Architecture
- hybrid Hybrid Gated DeltaNet + Gated Attention, MoE (512 experts, 10 routed + 1 shared), MTP
- Total parameters
- 2.4T sources
- Active parameters
- 95B
- Checkpoint parameters
- –
- Note
- Name indicates 2.4T total, 95B active
- Context length
- 262,144 tokens
- Input
- text
- Output
- text
Where this comes from
Parameter counts as the developer states them. Where the checkpoint holds a different total, both are shown (register A16).
Openness and Access
- Openness
- open weights restricted
- Licence
- Qwen3.8-Max License sources
- Access routes
- open weights download
Where this comes from
The licence file governs over registry tags and aggregator labels.
Deployment
- Runs on
- multi-GPU node
- Footprint
- FP8 variant
- Runtimes
- not in the notes
Where this comes from
Deployment tiers and runtimes as named in the notes. Footprint is the notes' figure, not a measurement by AURION.
Indonesia
- Indonesian support
- not stated
- Regional languages
- not in the notes
- Jakarta hosting
- unknown
- Languages
- not in the notes
Where this comes from
Indonesia fields are carried on every model; unknown is an allowed value.
Industry Use Cases
No industry use case links this model.
Links
Developer
Alibaba (Qwen)big tech, China
Alibaba (Qwen)big tech, ChinaModel
Qwen3.8-2.4T-A95BAlibaba (Qwen), reasoning
Qwen3.8-2.4T-A95BAlibaba (Qwen), reasoningR&D TopicsReasoning via Large-Scale RL
Use CasesChinese Medical LLMsHealthcareUse cases of models in the same R&D topic; none is recorded for this model itself.
Who Runs Itnone recordedDeployments of these use cases; none names this model.
Organisations 1
Sources
- huggingface.co/api/models/Qwen/Qwen3.8-2.4T-A95B https://huggingface.co/api/models/Qwen/Qwen3.8-2.4T-A95B
- huggingface.co/Qwen/Qwen3.8-2.4T-A95B/raw/main/README.md https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B/raw/main/README.md
- huggingface.co/Qwen/Qwen3.8-2.4T-A95B/raw/main/LICENSE https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B/raw/main/LICENSE
- Qwen3.8 2.4T A95B - Intelligence, Performance & Price Analysis | Artificial Analysis artificialanalysis.ai https://artificialanalysis.ai/models/qwen3-8-2-4t-a95b
- Qwen (Qwen) huggingface.co https://huggingface.co/Qwen
- Qwen/Qwen3.8-2.4T-A95B , Hugging Face huggingface.co https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B
- LICENSE , Qwen/Qwen3.8-2.4T-A95B at main huggingface.co https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B/blob/main/LICENSE
- Open-Source LLMs 2026: Which Are Actually Open-Licensed codersera.com https://codersera.com/blog/open-source-llms-landscape-2026/
- runtimewire.com/article/alibaba-qwen38-24t-open-weights-enter... https://runtimewire.com/article/alibaba-qwen38-24t-open-weights-enterprise-license
To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Qwen3.8-2.4T-A95B", https://aurion.semai.id/models/alibaba-qwen3-8-2-4t-a95b/, records of October 2026.