Project AURIONAI Use-Cases Research & Innovation
Alibaba (Qwen)
LLM

Qwen3.6-35B-A3B

General open model marketed as an agentic coder; the strongest small open agentic coder found in the research notes.

Benchmark bar charts for Qwen3.6-35B-A3B against other models.
Levelmodel
ClassLLM
Parameters35B
Context262,144
Released2 Apr 2026
Opennessopen weights permissive
Runs onconsumer GPU, laptop
Statusactive

Benchmark Results

vendor-reportedmaintainer-run
SEA-HELMSoutheast Asian and Indonesian
Checkpoint Qwen3.6-35B-A3B as listed; release view (data last updated 18 Sep 2026); score is the average of 2000 bootstrapsmaintainer-run
77.29 95% CI [-1.22, +1.21]
04080

Score in score. Higher is better.

SEA-HELM leaderboard, 58 models, Indonesian (ID)18 Sep 2026 leaderboard.sea-lion.ai
Checkpoint Qwen3.6-35B-A3B as listed; release view (data last updated 18 Sep 2026); score is the average of 2000 bootstrapsmaintainer-run
73.86 95% CI [-0.60, +0.57]
04080

Score in score. Higher is better.

SEA-HELM leaderboard, 58 models, SEA average18 Sep 2026 leaderboard.sea-lion.ai
vendor-reported
73.4
050100

Score in %. Higher is better.

Terminal-Bench 2.0agents and tool use
vendor-reported
51.5
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).

R&D Topics

Identity

Developer
Alibaba (Qwen)
Family
Qwen3.6
Released
2 Apr 2026 sourcessecondary report
Weights published
–
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
MoE
Total parameters
35B sources
Active parameters
3B
Checkpoint parameters
–
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 permissive
Licence
Apache License 2.0 sources
Access routes
open weights download
Where this comes from
The licence file governs over registry tags and aggregator labels.

Deployment

Runs on
consumer GPU, laptop
Footprint
T4/T5 estimate (unverified). About 1M with YaRN.
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.

Lineage

The notes record no lineage for this model.

Industry Use Cases

No industry use case links this model.

Sources

  1. Qwen3.6-35B-A3B: Alibaba Open-Sources a Frontier-Class Agentic Coder rits.shanghai.nyu.edu https://rits.shanghai.nyu.edu/ai/qwen3-6-35b-a3b-alibaba-open-sources-a-frontier-class-agentic-coder/
  2. Qwen/Qwen3.6-27B , Hugging Face huggingface.co https://huggingface.co/Qwen/Qwen3.6-27B

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Qwen3.6-35B-A3B", https://aurion.semai.id/models/alibaba-qwen3-6-35b-a3b/, records of October 2026.