QQwen · Released Apr 27, 2026
Qwen3.6 35B A3B
Qwen3.6-35B-A3B is an open-weight multimodal model from Alibaba Cloud with 35 billion total parameters and 3 billion active parameters per token. It uses a hybrid sparse mixture-of-experts architecture combining Gated...
- Top 47%OTIS Mock AIME
- Top 50%GPQA Diamond
- Released
- Apr 27 2026
- Context window
- 262K 236K out
- Input / 1M tokens
- $0.10 from Free
- Output / 1M tokens
- $0.95 from Free
Where to run it
45 providers compared
Free to try on InferX, Hetzner, NaN — free endpoints are often limited-time and rate-limited. Included in Umans AI Coding Plan. Cheapest: engy at $0.045 / $0.30 per 1M input / output tokens — 98% less than the priciest option.
| Provider | Input | Output | Cache read | Context | Max output | Uptime | vs. best |
|---|---|---|---|---|---|---|---|
| InferXfree · may be limited-time | Free | Free | — | 262K | 66K | — | Free |
| Hetznerfree · may be limited-time | Free | Free | Free | 262K | 262K | — | Free |
| NaNfree · may be limited-time | Free | Free | — | 262K | 66K | — | Free |
| Umans AI Coding Plansubscription plan | Plan | Plan | Plan | 262K | 262K | — | Plan |
| engy | $0.045 | $0.30 | $0.015 | 208K | 8K | — | Cheapest |
| Boundless | $0.04 | $0.35 | $0.02 | 262K | 66K | — | +8% |
| EmpirioLabs AI | $0.07 | $0.42 | $0.035 | 131K | 16K | — | +45% |
| DarkbloomFP4 | $0.05 | $0.70 | $0.025 | 262K | 33K | 100%24h | +95% |
| SaladCloud AI Gateway | $0.09 | $0.60 | — | 262K | 262K | — | +100% |
| Zenifra | $0.19 | $0.48 | — | 262K | 66K | — | +141% |
| Cortecs | $0.17 | $0.56 | — | 262K | 33K | — | +143% |
| AKI.IO | $0.17 | $0.56 | — | 256K | 33K | — | +145% |
| AkashMLFP8 | $0.10 | $0.90 | $0.05 | 262K | 236K | 100%24h | +176% |
| ParasailFP8 | $0.10 | $0.95 | $0.05 | 262K | 236K | 100%24h | +187% |
| DeepInfraFP8 | $0.10 | $0.95 | $0.10 | 262K | 66K | 98.9%24h | +187% |
| Kilo Gateway | $0.10 | $0.95 | $0.05 | 262K | 236K | — | +187% |
| VeniceFP8 | $0.10 | $1.00 | — | 256K | 66K | 100%24h | +199% |
| DekaLLM | $0.10 | $1.00 | $0.05 | 262K | 236K | 99.9%24h | +199% |
| Venice AI | $0.10 | $1.00 | — | 256K | 66K | — | +199% |
| Hugging Face | $0.15 | $0.95 | — | 262K | 66K | — | +222% |
| Pioneer | $0.14 | $1.00 | $0.028 | 262K | 131K | — | +226% |
| Umans AI | $0.15 | $1.00 | $0.05 | 262K | 33K | — | +233% |
| Ambient | $0.15 | $1.00 | $0.15 | 262K | 262K | — | +233% |
| NEAR AI Cloud | $0.17 | $1.10 | $0.056 | 262K | 8K | — | +270% |
| Phala | $0.20 | $1.27 | $0.056 | 262K | 236K | 96.4%24h | +330% |
| NanoGPTno tool calling | $0.20 | $1.27 | $0.10 | 262K | 262K | — | +330% |
| CoreWeaveFP8 · no tool calling | $0.25 | $1.25 | $0.25 | 262K | 236K | 100%24h | +360% |
| AtlasCloudFP8 | $0.25 | $1.25 | $0.25 | 262K | 66K | 99.4%24h | +360% |
| Neuralwatt | $0.29 | $1.15 | $0.029 | 131K | 131K | — | +364% |
| LLM Gateway | $0.25 | $1.49 | — | 262K | 64K | — | +412% |
| Novita AI | $0.25 | $1.49 | — | 262K | 66K | — | +412% |
| DevPass (LLM Gateway) | $0.25 | $1.49 | — | 262K | 66K | — | +412% |
| Merge Gateway | $0.25 | $1.49 | $0.05 | 262K | 66K | — | +412% |
| OrcaRouter | $0.25 | $1.49 | — | 262K | 66K | — | +412% |
| Opper | $0.25 | $1.49 | — | 262K | 66K | — | +412% |
| AIHubMix | $0.25 | $1.52 | — | 262K | 66K | — | +426% |
| evroc | $0.34 | $1.38 | — | 262K | 66K | — | +455% |
| SiliconFlow (China) | $0.27 | $1.61 | — | 262K | 66K | — | +457% |
| SiliconFlowFP8 · no tool calling | $0.24 | $1.80 | $0.15 | 262K | 236K | 96.6%24h | +479% |
| Scaleway | $0.28 | $1.68 | — | 128K | 16K | — | +480% |
| 302.AI | $0.28 | $1.70 | — | 262K | 66K | — | +484% |
| Thinking Machines | $0.45 | $1.33 | $0.09 | 66K | 66K | — | +517% |
| GreenPT | $0.34 | $2.05 | — | 262K | 33K | — | +608% |
| Alibaba | $0.38 | $2.25 | — | 262K | 66K | — | +676% |
| LLMTRno tool calling | $5.00 | $10.00 | — | 16K | 16K | — | +5647% |
Value
Price vs. performance
Qwen3.6 35B A3B Other Qwen Other labs Value frontier · GPQA Diamond vs. input price per 1M tokens (log scale) · Drag to pan, pinch or Ctrl + scroll to zoom, double-click to zoom in.
Benchmarks
8 benchmark scores
| Benchmark | Score | Rank | Leader |
|---|---|---|---|
| Terminal-BenchSoftware engineering · Tasks completed with the programs in a real terminal (best agent) | 24.6% | #34of 40 | GPT-5.584.7% |
| SciCodeSoftware engineering · Research-coding problems from the natural sciences | 36.6% | #86of 102 | Claude Opus 5.566.9% |
| WebDev ArenaDesign & visual · People vote on which model builds the better website (Elo) | 1250 | #84of 90 | Claude Opus 5.51820 |
| GPQA DiamondReasoning & knowledge · PhD-level biology, chemistry and physics questions | 84.9% | #61of 122 | Gemini 4 Argon96.4% |
| CritPtReasoning & knowledge · Open-ended, research-level physics problems | 0.3% | #88of 111 | GPT-5.6 Sol32.3% |
| LMCAReasoning & knowledge · Judge the quality of conceptual arguments like experts | 29.7% | #78of 123 | Claude Sonnet 5.568.3% |
| FrontierMathMaths · Expert-written problems, undergrad to early research (tiers 1–3) | 20.3% | #71of 87 | GPT-6.1 Sol93.7% |
| OTIS Mock AIMEMaths · Competition-style problems harder than MATH Level 5 | 86.7% | #56of 120 | Gemini 4 Argon100.0% |
Specs
Details
- Model ID
qwen/qwen3.6-35b-a3b- Developer
- Qwen
- Release date
- Apr 27, 2026 ·
- Context window
- 262,144 tokens
- Max output
- 235,929 tokens
- API price
- $0.10 / $0.95 per 1M input / output tokens
- Providers
- 45
- Input types
- text, image, video
- Reasoning
- Yes
- Tool calling
- Yes
- Open weights
- Yes
- Free tier
- No
Release date and prices from OpenRouter's catalog; providers from OpenRouter and models.dev. Buying directly from a provider can cost a little differently. Speed, latency and uptime are measured on OpenRouter only; uptime is the share of OpenRouter requests each provider served over the last day. Benchmark scores are the best result per benchmark from Epoch AI, licensed CC BY 4.0. Rank counts every model tracked here with a score.