Qwen3 VL 235B A22B Thinking
Qwen · released Sep 22, 2025 · Qwen/Qwen3-VL-235B-A22B-Thinking
- Type
- Open weightsApache License 2.0
- Params
- 236B
- Context
- 131K
22B active per word · about 98K words of context
Our take
Written Aug 2, 2026Qwen3 VL is a large multimodal model from Alibaba whose weights can be downloaded under a permissive Apache licence. It uses a mixture-of-experts design that keeps only 22 billion parameters active per token, and it scores highest on coding tasks while trailing on creative writing.
Choose this for vision-language work that needs open weights and a permissive licence, or when coding performance is the priority. It suits cost-sensitive inference where the cheapest tracked input rate matters, or throughput-sensitive deployment where the fastest measured host is key. Skip it if creative writing quality is central, or if you need consistent output pricing across providers.
The case for it
- Coding performance sits 59.2 points above its overall chat ranking on the Arena leaderboard.
- Apache 2.0 licence allows commercial use, fine-tuning and redistribution.
- 22 billion active parameters per token from 235.7 billion total — roughly tenfold parameter expansion.
- Lowest input pricing among tracked offers, well under half the next-cheapest host.
The case against it
- Creative writing is the weakest of its measured skills, 116.4 points below its own coding score and 57.2 points below its overall text ranking.
- Output pricing varies sharply even for the same model from the same provider, with no disclosed reason.
- The cheapest tracked offer lists no throughput figure; the measured speed comes from a pricier tier.
How good is it?
IntelligencePuzzles, maths, exam questions
Arena Text (overall)80th of 143 · 1395.3via Thinking
CodingWriting and fixing code on its own
Arena Coding71st of 143 · 1454.6via Thinking
Arena Coding is the only board that has scored it for this.
AgenticPlanning, calling tools, staying on task
Nobody we watch has scored Qwen3 VL 235B A22B Thinking for this. We would take the rating from Arena Agent (IPS).
WritingWe do not rate this
Two boards come close and neither tests writing: Arena Creative Writing asks people which of two replies they prefer, and LiveBench Language tests whether a model understood a passage. So we show where Qwen3 VL 235B A22B Thinking placed and give it no mark out of five.
These tests check whether a model follows instructions — a precondition for all the work above, but not a measure of how well that work is done, which is why they get no rating.
Every published score for this model6 scoresEvery figure we hold, from 6 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Can you run it yourself?
- Fits in memory
- weights load entirely on the card
- Spills to system RAM
- some weights offload; much slower
- Too large
- will not load even with offload
- est
- size is calculated; the verdict could change by 10%
GeForce RTX 4090 · 24 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Apple M1 Pro (16-core GPU) · 32 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Comfortable fit
Apple M3 Ultra (80-core GPU) · 512 GB
Room to spare. 229.3 GB spare means a 10% error in the size would not change the answer.
Memory use by level
Against a 24 GB card.
All 3 sizes here are calculated, not measured. We hold no measured file for this model, so each size comes from the parameter count and every verdict above inherits that uncertainty.
Check against your own machine → · All 71 devices, with every size →
Or rent it from someone else
Cheapest of 5 live listings. Picked at the widest standard context we hold, within one quantisation slice, so the numbers beside it are a price one host actually charges.
- per 1M tokens
- $0.26 in / $2.60 out
- Context served
- 131K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| Alibaba Cloud | $0.26 / $2.60 | 131K | not measured | No | Yesunknown period | Unknown |
| Novita AI | $0.98 / $3.95 | 131K | not measured | Unknown | Unknown | Unknown |
| Novita AIbf16 | $0.98 / $3.95 | 131K | 41 tok/s | No | No | Confirmed |
| OpenRouter | $0.40 / $4.00 | 131K | not measured | Unknown | Unknown | Unknown |
| Alibaba Cloudfp8 | $0.40 / $4.00 | 131K | 60 tok/s | No | Yesunknown period | Unknown |
Novita AI
OpenRouterAcross the 5 listings we hold: 3 say they do not train on prompts, 0 say they do and 2 do not say. 1 appear in the zero-retention registry we check; the rest are unknown to us rather than confirmed either way.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
- ✓
- Supported
- ✗
- Not supported
- Not published
- host gave no parameter list
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| Alibaba Cloud | |||
| Novita AI | |||
| Novita AIbf16 | ✓ | ✗ | ✗ |
| OpenRouter | ✓ | ✓ | ✓ |
| Alibaba Cloudfp8 | ✓ | ✓ | ✓ |
Tool calling: 3 of 5 listings say yes, 2 publish no parameter list. JSON output: 2 of 5 listings say yes, 1 says no, 2 publish no parameter list. Strict schema: 2 of 5 listings say yes, 1 says no, 2 publish no parameter list.
When we formed this view
Dates behind this page
Prices last checked 14h ago
What we do not know about this model yet
- We hold no measured file for it, so all 3 sizes on this page are calculated from the parameter count.
- 2 of 5 listings publish no parameter list, so what their API accepts is unknown to us.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 2 of 5 listings do not say whether they train on prompts.
- We hold no cached-input rate for any of its listings.
Licence and identifiers
What the licence allowsApache License 2.0, what it allows commercially, and the identifiers you need to pull this model — its Hugging Face repo, our slug and a machine-readable card.
Licence
Apache License 2.0
Fully permissive: commercial use, redistribution, and derivatives allowed. Requires attribution and a copy of the license. Includes an express patent grant.
Identifiers
- Hugging Face
- Qwen/Qwen3-VL-235B-A22B-Thinking
- Architecture
- Mixture of experts
- Modality record
- text+image->text
- Catalogue slug
- qwen-qwen3-vl-235b-a22b-thinking