Model reference · open weights

A3-Qwen3.5

A3-Qwen3.5 is an open-weight language model from McGill-NLP, listed in the AxForge catalogue. AxForge can bring it up on EU-owned hardware for you on request — with the licence handled where one is required.

Licence fee required LLMs McGill-NLP 3 variants 1k downloads/mo
Request a licence + hosting quote All served models Not on the shared API today — deployed on request.

About

What A3-Qwen3.5 is

A3-Qwen3.5-9B Structured Distillation of Web Agent Capabilities Enables Generalization Xing Han Lù, Siva Reddy A3-Qwen3.5-9B is a 9B multimodal web agent fine-tuned from Qwen/Qwen3.5-9B on A3-Synth, a synthetic dataset generated using the Agent-as-Annotators (A3) framework. The model achieves 41.5% on WebArena, surpassing closed-source models such as Claude 3.5 Sonnet (36.0%) and GPT-4o (31.5%) under the same evaluation protocol. Usage Serve with vLLM: Training - Base model: Qwen/Qwen3.5-9B - Data: A3-Synth (16k examples from Gemini 3 Pro trajectories) - Method: SFT with FSDP - Max sequence length: 16,384 - Learning rate: 1e-5 - Epochs: 2 - Batch size: 1 per GPU, gradient accumulation 4 Model Variants Citation

Summarised from the published model card. Read the full card on the HuggingFace links below.

Specifications

What it is

MakerMcGill-NLP
TypeLanguage models
Parameters (lead)9.4B
Variants3
Runs withtransformers
Based onQwen/Qwen3.5-9B
Released2026-04-06
Popularity1k downloads / month
Likes8
LicenceCommercial licence needed

How it works

How language models work

Your prompttext / messagesTransformerattention over tokensNext-token loopgenerate + streamResponsetext · tool callsA language model reads your tokens and predicts the next one, again and again, streaming the reply back.

Variants

Sizes & precisions

Open weights ship in several sizes and precisions. One page, all the variants — pick the one that fits your GPU. VRAM figures are estimates from model size.

VariantParamsPrecisionVRAMFits 16 GBWeights
A3-Qwen3.5-9B9.4BBF16~21.6 GBWeights ↗
A3-Qwen3.5-4B5.2BBF16~11.9 GBWeights ↗
A3-Qwen3.5-2B2.7BBF16~6.3 GBWeights ↗

Using it via the API

Call it like any OpenAI endpoint

Once AxForge deploys a3-qwen3-5 for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (a3-qwen3-5 below is illustrative; you get the exact model name on deployment.)

$ curl -sS https://api.axforge.ai/v1/chat/completions \
  -H "Authorization: Bearer $AXFORGE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"a3-qwen3-5","messages":[{"role":"user","content":"Hello"}]}'

Details

Languages, data & research

Languages

en

Tags

transformers safetensors qwen3_5 image-text-to-text agents web sft qwen conversational en endpoints_compatible

Papers

Licence

Commercial licence needed

The weights are open but its licence needs a commercial agreement for business use. AxForge can arrange that licence and host the model for you — you pay AxForge, we settle with the model’s maker. Ask us for a quote. Read the licence ↗

Sources

Weights & code

Want A3-Qwen3.5 on EU-owned hardware?

Request a licence + hosting quote See what’s served now

Explore

More language models

© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms