Qwen3.6 27B
Qwen · released Apr 21, 2026 · Qwen/Qwen3.6-27B
- Type
- Open weightsApache License 2.0
- Params
- 27.8B
- Context
- 262K
about 197K words of context
Our take
Written Aug 2, 2026Alibaba's 27-billion-parameter mid-size model can be downloaded, released in 2026. It accepts text, images and video and is a practical self-hosting sweet spot for teams that want modern downloadable capability without needing a server-grade GPU.
Use this for mid-size deployments balancing capability against GPU cost. The 27-billion-parameter class is a comfortable self-hosting sweet spot. Pick it for multimodal input without big-model prices, or as a modern fine-tuning base. Skip it if you need measured quality scores or the lowest output price in the class.
The case for it
- 27 billion parameters: a standard compressed size fits in 24GB-class GPUs, unlike the 400-billion-plus open mixture-of-experts models.
- Text, image and video input in a mid-size model.
The case against it
- No benchmark scores yet, so there is no measured quality data.
- Output price is relatively high for its size class.
How good is it?
IntelligencePuzzles, maths, exam questions
Qwen3.6 27B is not on Arena Text (overall), which is where the rating would come from, so there is no rating here. It is on LiveBench Data Analysis, in 24th of 35 with 70.4.
CodingWriting and fixing code on its own
Qwen3.6 27B is not on Arena Coding, which is where the rating would come from, so there is no rating here. It is on LiveBench Coding, in 26th of 35 with 71.8.
AgenticPlanning, calling tools, staying on task
Qwen3.6 27B is not on Arena Agent (IPS), which is where the rating would come from, so there is no rating here. It is on LiveBench Agentic Coding, in 33rd of 35 with 39.3.
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 Qwen3.6 27B 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 model8 scoresEvery figure we hold, from 8 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%
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.
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 13 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
- $0.29 in / $2.40 out
- Context served
- 262K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| Io Netfp8 | $0.27 / $1.89 | 33K | 34 tok/s | No | No | Confirmed |
| Chutesfp8 | $0.30 / $2.00 | 262K | 30 tok/s | No | Yesunknown period | Unknown |
| Morph | $0.29 / $2.40 | 131K | 25 tok/s | No | No | Confirmed |
| OpenRouter | $0.29 / $2.40 | 262K | not measured | Unknown | Unknown | Unknown |
| Phala | $0.32 / $2.70 | 262K | 51 tok/s | No | No | Confirmed |
| Alibaba Cloudfp8 | $0.45 / $2.70 | 262K | 57 tok/s | No | Yesunknown period | Unknown |
| Alibaba Cloud | $0.45 / $2.70 | 262K | 33 tok/s | No | Yesunknown period | Unknown |
| DeepInfrafp8 | $0.32 / $3.20 | 262K | not measured | Unknown | Unknown | Unknown |
| SiliconFlowfp8 | $0.30 / $3.20 | 262K | 34 tok/s | No | No | Confirmed |
| DeepInfrafp8 | $0.32 / $3.20 | 262K | 59 tok/s | No | No | Confirmed |
| Venice AIfp8 | $0.33 / $3.25 | 256K | 69 tok/s | No | No | Confirmed |
| Novita AI | $0.60 / $3.60 | 262K | not measured | Unknown | Unknown | Unknown |
| CoreWeavefp8 | $0.60 / $3.60 | 262K | 92 tok/s | No | No | Confirmed |
Morph
DeepInfra
Novita AIAcross the 13 listings we hold: 10 say they do not train on prompts, 0 say they do and 3 do not say. 7 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 |
|---|---|---|---|
| Io Netfp8 | ✓ | ✓ | ✗ |
| Chutesfp8 | ✓ | ✗ | ✓ |
| Morph | ✓ | ✓ | ✓ |
| OpenRouter | ✓ | ✓ | ✓ |
| Phala | ✓ | ✓ | ✗ |
| Alibaba Cloudfp8 | ✓ | ✓ | ✓ |
| Alibaba Cloud | ✓ | ✓ | ✓ |
| DeepInfrafp8 | |||
| SiliconFlowfp8 | ✓ | ✓ | ✓ |
| DeepInfrafp8 | ✓ | ✓ | ✓ |
| Venice AIfp8 | ✓ | ✓ | ✓ |
| Novita AI | |||
| CoreWeavefp8 | ✓ | ✓ | ✓ |
Tool calling: 11 of 13 listings say yes, 2 publish no parameter list. JSON output: 10 of 13 listings say yes, 1 says no, 2 publish no parameter list. Strict schema: 9 of 13 listings say yes, 2 say no, 2 publish no parameter list.
Models people weigh against Qwen3.6 27B
When we formed this view
Dates behind this page
Prices last checked 35h 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 13 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.
- 3 of 13 listings do not say whether they train on prompts.
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/Qwen3.6-27B
- Architecture
- Dense
- Modality record
- text+image+video->text
- Catalogue slug
- qwen-qwen3-6-27b