See how ChatGPT, Perplexity and Google AI Overviews describe you today

Free AI Visibility Audit
Qwen model°

Qwen3.6 35B A3B

Qwen3.6 35B A3B is an open-weight multimodal model from Qwen, Alibaba Cloud's model family, that reads text, images and video and activates only a small fraction of its parameters on each token.

QwenVerified 25/09/2026Released 27/04/2026

01 / decision

Decision snapshot

Each figure sits next to the middle of the field, so it can be read as dear or cheap, wide or narrow, rather than floating on its own. Compared against the 231 models we track, not against an absolute standard.

Capability

Not measured

No published benchmark scores yet

Input / 1M tokens

$0.15

field median $0.42

Output / 1M tokens

$1.00

field median $1.81

Context

262K

field median 500K

02 / overview

What Qwen3.6 35B A3B is, and when to reach for it

Four questions, answered separately, because somebody arrives at one of them rather than at the top of the page.

What it is

The job it was built for, and the job it is not for.

Qwen3.6 35B A3B is a general-purpose multimodal model built for high-volume work: reading documents, screenshots and video alongside text, and returning long-form output. Because the weights are open, it is meant to be run by the team that uses it as much as it is called through a hosted endpoint. It is not a specialist reasoning or coding flagship, and it is not the model to reach for when you need the single strongest answer regardless of cost.

When it arrived, and when to use it

Where it sits in its line, and when a sibling is the better pick.

Qwen3.6 35B A3B first appeared in April 2026, as the small, sparsely activated member of the Qwen3.6 line rather than its headline model. It is the right pick when you are running a large number of calls, need image or video input, or need to host the model yourself. It is the wrong pick when a task needs the deepest reasoning a frontier model can give, or when you want a single vendor to carry the operational burden.

How you reach it

The API, the apps it powers, and what its limits let you do.

Qwen3.6 35B A3B is reached through Alibaba Cloud's API and through the third-party inference providers that host open Qwen weights, and it can also be downloaded and run on your own hardware. The context window is large enough to hold a full documentation set, a long transcript or a video alongside the instructions about it, and the output ceiling is unusually high, so a single call can produce a complete long report or a bulk translation rather than a fragment you stitch together.

Why it matters

What changes because this exists, or why it does not.

Qwen3.6 35B A3B makes it practical to put multimodal processing on the cheap, repetitive end of a pipeline: classifying screenshots, summarising video, extracting fields from scanned documents at volume. Work that previously had to be routed to a costly frontier model, or split into many small calls because of output limits, can now sit in one place. For teams with a compliance or data residency constraint, the open weights mean the same capability can run inside their own environment.

Follows Qwen3.6 27B. Superseded by Qwen3.6 Flash. See the whole line.

03 / evidence

How much of this is verified

Split by category, so a strong number never hides a thin evidence base. Verified means we read it on the benchmark's own published results; a provider's claim about its own model is shown and labelled rather than dropped.

No published benchmark scores for this model yet.

We publish a score only where we can link the result to where it was published. Until a benchmark result for this model exists in a source we read, this section stays empty rather than being filled with a provider's marketing figure.

How we decide what counts as evidence

04 / ledger

Benchmark ledger

Every published row, grouped by category, each compared with the best published score on the same benchmark. 'Is 64% good' is a question nobody can answer; '26 points behind the leader' is one anybody can.

Nothing in the ledger yet.

Each row here carries a score, the benchmark it came from, what the leading model scored on the same test, and a link to the published result. Rows appear as results are published and read.

How we decide what counts as evidence

05 / capability

Capability shape

Where this model is strong, and against how many peers. Ranks are against models with evidence in that category, not against everything we track: ranking against models nobody tested would rank who published, not who is better.

No category scores to shape yet.

A category score is the weighted mean of the benchmarks published for it. With no published rows there is nothing to average, and an empty chart drawn at zero would say something false.

How we decide what counts as evidence

06 / cost

What it costs

List API rates as last read from the provider, with the source on every row, plus every change we have recorded since we started tracking it.

Qwen3.6 35B A3B API pricingSurge45°
ChargePriceUnitRead onSource
Input$0.15per 1M tokens2026-09-24Check
Output$1.00per 1M tokens2026-09-24Check

Qwen3.6 35B A3B sits at the low end of the market on both input and output, cheap enough that cost stops being the thing that shapes your architecture. Output is priced well above input, as is normal, so the long generations its output ceiling allows are where the bill actually accumulates.

What a month costsSurge45°
WorkloadInput / monthOutput / monthCost
A small product team20M tokens5M tokens$8.00
A busy support assistant200M tokens40M tokens$70.00
A document pipeline1000M tokens100M tokens$250.00

List API rates, no caching and no batch discount, which both providers offer and which change the answer a great deal. Treat these as the ceiling, not the bill.

07 / specs

Specifications

As listed by the provider's own catalogue and re-read every few hours. Anything absent is absent there too.

SpecificationSurge45°
Context window262,144 tokens
Maximum output235,929 tokens
Modalitiestext, image, video
Released27/04/2026
StatusCurrent
Catalogue identifierqwen/qwen3.6-35b-a3b

08 / lineage

Lineage

What this model replaced, what replaced it, and what else its provider has in the field.

Also from Qwen

09 / line

The line

Every model in this family in release order, so a page from eight months ago says in one glance that two newer ones exist.

  1. 01Qwen3 14B28/04/2025
  2. 02Qwen3 235B A22B28/04/2025
  3. 03Qwen3 30B A3B28/04/2025
  4. 04Qwen3 32B28/04/2025
  5. 05Qwen3 8B28/04/2025
  6. 06Qwen3.5 397B A17B16/02/2026
  7. 07Qwen3.5 Plus 2026-02-1516/02/2026
  8. 08Qwen3.5-122B-A10B25/02/2026
  9. 09Qwen3.5-27B25/02/2026
  10. 10Qwen3.5-35B-A3B25/02/2026
  11. 11Qwen3.5-Flash25/02/2026
  12. 12Qwen3.5-9B10/03/2026
  13. 13Qwen3.6 Plus02/04/2026
  14. 14Qwen3.5 Plus 2026-04-2027/04/2026
  15. 15Qwen3.6 27B27/04/2026
  16. 16Qwen3.6 35B A3B27/04/2026
  17. 17Qwen3.6 Flash27/04/2026
  18. 18Qwen3.6 Max Preview27/04/2026
  19. 19Qwen3.7 Max21/05/2026
  20. 20Qwen3.7 Plus03/06/2026
  21. 21Qwen3.7 Flash27/07/2026
  22. 22Qwen3.8 2.4T A95B12/08/2026
  23. 23Qwen3.8 27B14/08/2026
  24. 24Qwen3.8 Flash26/08/2026
  25. 25Qwen3.8 Max (0902)03/09/2026
  26. 26Qwen3.8 Omni Flash21/09/2026
  27. 27Qwen3.8 Max Prime23/09/2026

Ordered by release date and worked out from the naming, so a new member slots in as soon as its page exists. A retired model keeps its page and its place in the line.

10 / notes

Our notes

What this model changes for a brand trying to be cited in AI answers, and every change we have logged since it launched.

What it changes for you

A model this cheap to run gets used for the bulk, unglamorous passes over content: crawling, summarising, classifying, extracting. That means more of the machine reading of your site and your documentation will be done by something that also reads images and video, so diagrams, screenshots and product demos are now part of what an AI engine can actually understand about you. Open weights also mean this reading may happen inside a customer's own infrastructure, where you get no log line and no referral, which is one more reason to judge visibility by how you are described in answers rather than by traffic.

Where buyers meet this model

Buyers meet Qwen3.6 35B A3B mostly through the API, either from Alibaba Cloud directly or via the inference hosts and open-weight aggregators that serve Qwen models. It also turns up inside products that never name it, because its low running cost makes it a common default for the background multimodal steps in someone else's application. Consumer exposure is indirect, through Alibaba's own assistant surfaces and through third-party apps built on the open weights.

Change log

Nothing published here yet. Changes appear within hours of a provider announcing them.

11 / questions

Questions

The things people ask about this model, answered from what is on this page rather than from anywhere else.

Can Qwen3.6 35B A3B read images and video?
Yes. It accepts text, images and video as input and replies in text, so it can handle screenshots, scanned documents and video clips in the same call as the instructions about them.
Is Qwen3.6 35B A3B open-weight?
Yes. The weights are published, so it can be downloaded and run on your own hardware as well as called through Alibaba Cloud's API or a third-party inference host.
When should I use something stronger?
When the task needs the deepest reasoning available, or when a single answer matters more than the cost of producing it. Qwen3.6 35B A3B is built for volume, breadth of input types and long output, not for being the strongest model in the room.
Surge45°

Is Qwen3.6 35B A3B recommending you?

Models change what gets cited. We measure whether AI answers name your brand or your competitors across every assistant, and show you what to change.