Qwen3.6 Plus
Qwen · released Apr 2, 2026
Qwen 3.6 Plus builds on a hybrid architecture that combines efficient linear attention with sparse mixture-of-experts routing, enabling strong scalability and high-performance inference. Compared to the 3.5 series, it delivers...
Specification
- Context window
- 1M
- Max output
- 66K
- Knowledge cutoff
- Not stated
- Parameters
- Undisclosed
- Licence
- Proprietary
- Serving providers
- 1
- Moderated
- No
- Uptime
- 100.0%
Intelligence
40.5
88th percentile
Coding
54.5
Coding Index
Agentic
29.0
Agentic Index
Output speed
39 t/s
Median across providers
Latency
982ms
Time to first token
Cost per task
$0.17
Estimated
Benchmarks
Where the score comes from
The Intelligence Index is a composite. These are the underlying evaluations this model was actually measured on.
Evaluation scores
Percentage correct · higher is better
- τ²-bench (Telecom)97.7%
- GPQA Diamond88.2%
- IFBench75.2%
- AA-LCR (long context)72.3%
- Terminal-Bench Hard43.9%
- SciCode40.7%
- Humanity's Last Exam27.8%
An evaluation missing from this list was not run for this model — it is not a zero.
View as table
| Evaluation | Score |
|---|---|
| τ²-bench (Telecom) | 97.7% |
| GPQA Diamond | 88.2% |
| IFBench | 75.2% |
| AA-LCR (long context) | 72.3% |
| Terminal-Bench Hard | 43.9% |
| SciCode | 40.7% |
| Humanity's Last Exam | 27.8% |
Against its peers
Intelligence Index · this model highlighted, nearest peers in grey
- Inkling SmallThinking Machines41.2
- GPT-5.2-CodexOpenAI41.2
- 41.1
- GLM 5.1Z.ai41.0
- GPT-5.4 MiniOpenAI40.9
- GLM 5Z.ai40.6
- Qwen3.6 PlusQwen40.5
- GPT-5.4 NanoOpenAI39.7
Peers are the models sitting closest on the Intelligence Index — the set you would realistically choose between.
View as table
| Model | Intelligence |
|---|---|
| Inkling Small | 41.2 |
| GPT-5.2-Codex | 41.2 |
| Qwen3.6 Max Preview | 41.1 |
| GLM 5.1 | 41.0 |
| GPT-5.4 Mini | 40.9 |
| GLM 5 | 40.6 |
| Qwen3.6 Plus | 40.5 |
| GPT-5.4 Nano | 39.7 |
Percentile among all indexed models
Pricing
What it costs to run
List prices per million tokens, plus what one representative task works out to.
List price
- Input / 1M tokens
- $0.325
- Output / 1M tokens
- $1.95
- Cached input / 1M
- Not offered
- Blended 3:1
- $0.731
One task, estimated
$0.17
- Input tokens
- 50,000
- Output tokens
- 80,000
- Profile
- Reasoning
Estimated from list pricing: 50K input tokens plus 80K output tokens for reasoning models (25K for non-reasoning).
Serving providers
Speed and latency figures are medians across these providers, so a widely-served model reports a blend rather than any single endpoint.
Arena
Head-to-head generation quality
Elo from pairwise judgements, broken out by the kind of thing the model was asked to build.
- Overall Elo
- 1,271
- Win rate
- 52.6%
- Strongest at
- UI components
- Tournaments
- 3,520
Elo by category
Dot position on a 1,200–1,275 scale · Elo has no meaningful zero
- 1,271
- 1,261
- 1,256
- 1,256
- 1,246
- 1,208
Agent categories (full-stack apps, mobile apps) are only scored for models tested in agent harnesses.
View as table
| Category | Elo | Win rate |
|---|---|---|
| UI components | 1271 | 52.6% |
| Websites | 1261 | 52.5% |
| Data visualisation | 1256 | 51.6% |
| 3D scenes | 1256 | 51.6% |
| Game development | 1246 | 50.8% |
| SVG | 1208 | 51.6% |