Qwen

Qwen 3.6 35B-A3B

Name: Qwen 3.6 35B-A3B
Author: Qwen

Qwen 3.6 35B-A3B is a Qwen model with 262K context, Not verified in / Not verified out per million tokens, last verified October 25, 2026. Open weights: yes. Status: Legacy. Public-board scores are linked to the publisher; vendor-only figures stay in the watch-out, not on the board.

This profile refers to Alibaba's downloadable Qwen3.6-35B-A3B checkpoint with 262,144-token native context. Hosted Max, Plus and Flash variants are separate models.

Recorded profile verification October 25, 2026Individual provider fields may change

AI model specification details
Specification	Qwen 3.6 35B-A3B
Provider	Qwen
Tier	Balanced
Context window	262K
Max output	Not verifiedUnverified
Input / 1M tokens	Not verifiedUnverified
Output / 1M tokens	Not verifiedUnverified
Weights	Open
Parameters	35B total / 3B active (MoE)
Reasoning levels	Not verifiedUnverified
Modalities	text, image, video
License	Apache 2.0
Released	Not verifiedUnverified

Pricing tiers: Apache-2.0 Qwen3.6-35B-A3B native context is 262,144. A separate Model Studio qwen3.6-35b-a3b row lists International USD/MTok $0.375 input/$2.25 output for both modes, and Beijing $0.248/$1.485. These are named hosted-route quotes, not a universal native-weight price. Hosted Max Preview, Plus and Flash remain separate identities. No hard native output cap is established.

Best for

Open-weight deployments
Multilingual work

Watch out

Native thinking on/off and preserved thinking are supported; named effort menu and hard output cap remain unverified. The card recommends 32,768 generation tokens for most work or 81,920 for difficult benchmarks; neither is a product cap. Native 262,144 context can extend to 1,010,000 with runtime scaling. April 16 is a hosted listing, not a verified weight-publication day. Hosted family Legacy labels do not establish retirement of these Apache weights.

Source receipts

Catalog figures for this model were checked against the following sources.

Current scoped source review: exact publisher model, tariff, licence or lifecycle evidence (accessed 2026-10-25)
Current scoped source review: exact publisher model, tariff, licence or lifecycle evidence (accessed 2026-10-25)
Current scoped source review: exact publisher model, tariff, licence or lifecycle evidence (accessed 2026-10-25)
Current scoped source review: exact publisher model, tariff, licence or lifecycle evidence (accessed 2026-10-25)
Current scoped source review: exact publisher model, tariff, licence or lifecycle evidence (accessed 2026-10-25)
Current scoped source review: exact publisher model, tariff, licence or lifecycle evidence (accessed 2026-10-25)
Current scoped source review: exact publisher model, tariff, licence or lifecycle evidence (accessed 2026-10-25)

Common questions

Qwen 3.6 35B-A3B

Answered from the verified figures on this page rather than general guidance.

What is Qwen 3.6 35B-A3B's context window?

Qwen 3.6 35B-A3B accepts about 262K tokens of context. That only matters if you routinely send very long documents, large codebases, or multi-turn histories that approach that limit.

What is Qwen 3.6 35B-A3B best for?

Qwen 3.6 35B-A3B is a balanced tier from Qwen. It suits open-weight deployments, multilingual work. Native thinking on/off and preserved thinking are supported; named effort menu and hard output cap remain unverified. The card recommends 32,768 generation tokens for most work or 81,920 for difficult benchmarks; neither is a product cap. Native 262,144 context can extend to 1,010,000 with runtime scaling. April 16 is a hosted listing, not a verified weight-publication day. Hosted family Legacy labels do not establish retirement of these Apache weights.

Can I self-host Qwen 3.6 35B-A3B?

Qwen 3.6 35B-A3B publishes open weights, but self-hosting depends on the licence, hardware footprint, quantisation quality, and serving stack. A hosted API is often cheaper until you have measured throughput and concurrency on your own hardware.

Model comparison catalog Benchmark library