Qwen3.5-27B
Qwen · released Feb 24, 2026 · Qwen/Qwen3.5-27B
- Type
- Open weightsApache License 2.0
- Params
- 27.8B
- Context
- 262K
about 197K words of context
Our take
Written Sep 7, 2026Qwen3.5-27B is a 27.8-billion-parameter multimodal model from Alibaba with a permissive Apache licence and a 262,144-token request limit. Its measured coding skill sits well above its general text level, making it a natural fit for code-heavy workloads rather than pure creative writing.
Pick this when you need an Apache-licensed open-weight model for commercial use or redistribution, or for coding tasks where its highest measured score applies. It also suits long-context work up to 262,144 tokens and budget-conscious hosted inference. Skip it if creative writing or web development code are your main needs, or if you need verified active-parameter efficiency data.
The case for it
- Coding is its standout skill: 41.8 points above its general text score on the human-preference leaderboard.
- Apache 2.0 licence allows commercial use, fine-tuning and redistribution without restriction.
- 262,144-token request limit is unusually large for a model of this size.
- Seven measured sub-skills on the Arena leaderboard, with no major gaps in coverage.
The case against it
- Creative writing and web development code are more than 50 points below its coding peak — its two lowest measured skills.
- Throughput at the cheapest verified price is only 2 tokens per second; faster options cost more.
- Active parameter count is undisclosed, so efficiency claims remain unverified.
How good is it?
EverydayGeneral questions and everyday reasoning
Arena Text (overall)92nd of 168 · 1409
CodingWriting and fixing code on its own
Arena Coding93rd of 168 · 1451
AgenticPlanning, calling tools, staying on task
Not yet scored on Arena Agent.
WritingDrafting and rewriting prose
Arena Creative Writing96th of 168 · 1358
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 model7 scoresEvery figure we hold, from 7 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?
Comfortable fit
GeForce RTX 4090 · 24 GB
Room to spare. 3 GB spare means a 10% error in the size would not change the answer.
GeForce RTX 5090 · 32 GB
Room to spare. 11 GB spare means a 10% error in the size would not change the answer.
Apple M1 Pro (16-core GPU) · 32 GB
Room to spare. 4.2 GB spare means a 10% error in the size would not 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 4 hours and 10 hours ago — each listing carries its own date.
- per 1M tokens
- $0.20 in / $1.56 out
- Context served
- 262K
- Throughput
- ~35 tok/s
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| Alibaba CloudThrough OpenRouter | $0.20 / $1.56checked 4 hours ago | 262K66K max reply | 35 tok/s | No | Yesunknown period | Unknown |
| OpenRouterOpenRouter's own listing | $0.20 / $1.56checked 4 hours ago | 262K | not measured | Unknown | Unknown | Unknown |
| SiliconFlowfp8Through OpenRouter | $0.25 / $2.00checked 10 hours ago | 262K236K max reply | 10 tok/s | No | No | Confirmed |
| AtlasCloudfp8Through OpenRouter | $0.27 / $2.16checked 4 hours ago | 262K66K max reply | 26 tok/s | No | Yesunknown period | Unknown |
| PhalaThrough OpenRouter | $0.30 / $2.40checked 10 hours ago | 262K66K max reply | 11 tok/s | No | No | Confirmed |
| Novita AIbf16Direct and through OpenRouter | $0.30 / $2.40checked 4 hours ago | 262K66K max reply through OpenRouter | 40 tok/sthrough OpenRouter | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterConfirmed |
| DeepInfrafp8Direct and through OpenRouter | $0.26 / $2.60checked 4 hours ago | 262K82K max reply through OpenRouter | 55 tok/sthrough OpenRouter | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterConfirmed |
OpenRouter
Novita AI
DeepInfraAcross the 7 listings we hold: 6 say they do not train on prompts (2 of them only through OpenRouter), 0 say they do and 1 does not say. 4 appear in the zero-retention registry we check (2 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 |
|---|---|---|---|
| Alibaba CloudThrough OpenRouter | ✓ | ✓ | ✓ |
| OpenRouterOpenRouter's own listing | ✓ | ✓ | ✓ |
| SiliconFlowfp8Through OpenRouter | ✓ | ✓ | ✓ |
| AtlasCloudfp8Through OpenRouter | ✓ | ✓ | ✓ |
| PhalaThrough OpenRouter | ✗ | ✓ | ✓ |
| Novita AIbf16Direct and through OpenRouter | ✓ | ✓ | ✗ |
| DeepInfrafp8Direct and through OpenRouter | ✓ | ✓ | ✓ |
Tool calling: 6 of 7 listings say yes, 1 says no. JSON output: 7 of 7 listings say yes. Strict schema: 6 of 7 listings say yes, 1 says no.
Models people weigh against Qwen3.5-27B
When we formed this view
Recent changes
What moved
cache read −80% ($0.150 → $0.030 per 1M tokens)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.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 1 of 7 listings does not say whether it trains on prompts, and 2 answer only through OpenRouter, not for their own listing.
- 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/Qwen3.5-27B
- Architecture
- Dense
- Takes in, gives back
- Text, images and video in, text out
- Catalogue slug
- qwen-qwen3-5-27b