- Type
- Open weightsApache License 2.0
- Params
- 32.8B
- Context
- 131K
about 98K words of context
Our take
Written Sep 2, 2026Qwen3 is a 32.8-billion-parameter text model released in 2025 under a permissive Apache licence. Its measured coding skill sits well above its general chat score, and it runs from budget hosts up to a premium speed tier.
Pick this for Apache-licensed self-hosting or budget inference among mid-size open models. Choose the premium tier if you need 220 tokens per second and can justify the higher rate. Skip it if creative writing quality matters most, or if you need image or audio input.
The case for it
- Coding is its standout skill: 60 points above its own general chat score on the Arena leaderboard.
- Arena scores span a narrow 103-point range with no collapse — maths and hard prompts both clear 1360.
- Apache 2.0 licence allows commercial use, fine-tuning and redistribution without restriction.
The case against it
- Creative writing is the lowest of its six measured skills, 43 points below its overall chat score.
- Budget hosts offer 11–25 tokens per second; the 220 tokens per second tier costs several times more.
- Dense 32.8 billion parameters with no disclosed active-parameter count — no per-token efficiency claim possible.
How good is it?
An open text model for everyday questions and code, though drafting and prose are not its strong suit.
- drafts, rewrites and editingArena Creative Writing · 127th of 168
EverydayGeneral questions and everyday reasoning
Arena Text (overall)124th of 168 · 1347
CodingWriting and fixing code on its own
Arena Coding122nd of 168 · 1406
Arena Coding is the only board that has scored it for this.
AgenticPlanning, calling tools, staying on task
Not yet scored on Arena Agent.
WritingDrafting and rewriting prose
Arena Creative Writing127th of 168 · 1304
Arena Creative Writing is the only board that has scored it for this.
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.
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?
GeForce RTX 4090 · 24 GB
Loads, but do not expect an assistant. Some of the weights sit in ordinary system memory, which is far slower than the card.20.7 GB of weights, plus 2.4 GB for the software that runs it and the smallest conversation it can hold, comes to 23.1 GB against the 22.8 GB this 24 GB device leaves free.
Comfortable fit
GeForce RTX 5090 · 32 GB
Room to spare. 7.8 GB spare means a 10% error in the size would not change the answer.
Apple M1 Pro (16-core GPU) · 32 GB
Borderline fit on an estimated size. It leaves 1 GB spare on a size we calculated rather than measured, and a 10% error either way would change the answer.
Memory use by level
Against a 24 GB card.
What is quantisation? →This model on every device we track71 devicesThe Q4 build most people download, on each device: what the weights come to, how much context the memory leaves, and whether it runs. Smallest device that runs it first.
Check against your own machine → · Where to rent it hosted →
Or rent it from someone else
Prices checked between 2 hours and 21 days ago — each listing carries its own date.
- per 1M tokens
- $0.080 in / $0.28 out
- Context served
- 131K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenRouterOpenRouter's own listing | $0.080 / $0.28checked 2 hours ago | 131K | not measured | Unknown | Unknown | Unknown |
| DeepInfrafp8Direct and through OpenRouter | $0.080 / $0.28checked 2 hours ago | 41K16K max reply through OpenRouter | 20 tok/sthrough OpenRouter | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterConfirmed |
| Novita AIDirect | $0.10 / $0.45checked 21 days ago | 41K | not measured | Unknown | Unknown | Unknown |
| SiliconFlowfp8Through OpenRouter | $0.14 / $0.57checked 2 hours ago | 131K118K max reply | 11 tok/s | No | No | Confirmed |
DeepInfra
Novita AIAcross the 4 listings we hold: 2 say they do not train on prompts (1 of them only through OpenRouter), 0 say they do and 2 do not say. 2 appear in the zero-retention registry we check (1 of them only through OpenRouter); the rest are unknown to us.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| OpenRouterOpenRouter's own listing | ✓ | ✓ | ✓ |
| DeepInfrafp8Direct and through OpenRouter | ✓ | ✓ | ✓ |
| Novita AIDirect | |||
| SiliconFlowfp8Through OpenRouter | ✓ | ✓ | ✓ |
Tool calling: 3 of 4 listings say yes, 1 publishes no parameter list. JSON output: 3 of 4 listings say yes, 1 publishes no parameter list. Strict schema: 3 of 4 listings say yes, 1 publishes no parameter list.
Models people weigh against Qwen3 32B
When we formed this view
Recent changes
What moved
first indexed by our pipelineEach date is the day we first saw the change, or the day the maker announced it.
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.
- 1 of 4 listings publishes no parameter list, so what its API accepts is unknown to us.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 2 of 4 listings do not say whether they train on prompts, and 1 answers only through OpenRouter, not for its own listing.
- We hold no cached-input rate for any of its listings.
- We hold no batch or off-peak rate for any of its listings.
- We hold a decode speed for it, but no prompt-processing (prefill) figure, so how long the input side of a job takes is unknown to us.
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/Qwen3-32B
- Architecture
- Dense
- Takes in, gives back
- Text in, text out
- Catalogue slug
- qwen-qwen3-32b