QwQ: pricing, context window and benchmarks
QwQ is a multimodal model from Alibaba Cloud. It offers a 128,000-token context window, costs $4.57 per million input tokens and $16.94 per million output tokens, and averages 78.6 across the index's benchmark suite.
Pricing
QwQ costs $4.57 per million input tokens and $16.94 per million output tokens. A blended 3:1 input-to-output mix works out to about $7.66 per million tokens. Cached input is listed at $0.44 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $4.57 | per 1M tokens |
| Output | $16.94 | per 1M tokens |
| Cached input | $0.44 | per 1M tokens |
| Blended 3:1 | $7.66 | estimate per 1M tokens |
| 1M in + 1M out | $21.51 | estimate |
Benchmarks
Scores below are modelled estimates on a 0–100 scale. Use them to compare relative position, not as official vendor results.
| Benchmark | Score | Rank | Percentile | Relative |
|---|---|---|---|---|
| MMLU-Pro | 78.3 | 234 | 61th | |
| GPQA Diamond | 76.7 | 108 | 82th | |
| SWE-bench Verified | 68.3 | 138 | 77th | |
| HumanEval | 84.4 | 126 | 79th | |
| MATH-500 | 80.0 | 208 | 65th | |
| AIME 2025 | 78.8 | 104 | 83th | |
| LiveCodeBench | 71.2 | 164 | 73th | |
| IFEval | 84.3 | 207 | 66th | |
| TruthfulQA | 79.9 | 184 | 69th | |
| HellaSwag | 88.0 | 141 | 77th | |
| MMMU | 75.0 | 163 | 73th | |
| Arena Elo | 1415.0 | 167 | 72th |
What QwQ is good for
QwQ is tuned toward coding work. It is a reasonable candidate for code generation, review, debugging and repository-scale tasks, especially when paired with a retrieval layer. At $4.57 per million input tokens, it sits in the mid-range of the index, and the 128,000-token context window is enough for large files and multi-file diffs.
Strengths
- Large 128,000-token context window for long documents and agent traces.
- Open weights make self-hosting, fine-tuning and data-residency review possible.
- Accepts more than text, including image input where the provider supports it.
- Coding benchmarks are a relative strength within its tier.
Trade-offs
- No standout weakness in the indexed specifications, but real-world behaviour still needs testing.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| QwQ | Alibaba Cloud | 128000 | $4.57 | 78.6 |
| GLM-4.6 Base | Zhipu AI | 128000 | $4.53 | 78.5 |
| Reka Core Reasoning | Reka AI | 32768 | $13.11 | 78.5 |
| Magistral Instruct | Mistral AI | 65536 | $4.48 | 78.4 |
| Nova Pro Turbo | Amazon | 128000 | $13.18 | 78.4 |
| Qwen3-Coder Chat | Alibaba Cloud | 128000 | $4.59 | 78.4 |
FAQ
How much does QwQ cost?
QwQ is listed at $4.57 per million input tokens and $16.94 per million output tokens. Cached input is $0.44 per million tokens. A blended 3:1 input-to-output mix is about $7.66 per million tokens.
What is the context window of QwQ?
QwQ has a context window of 128,000 tokens and can emit up to 8,192 output tokens per request in this index.
Is QwQ open weights?
Yes. QwQ is listed as an open-weight model under a open weights licence.
How does QwQ score on benchmarks?
QwQ averages 78.6 across the scored benchmarks in this index, ranking #154 of 600 models. Its strongest relative result is Arena Elo at 1415.0.
What is QwQ best for?
QwQ is tuned toward coding work. It is a reasonable candidate for code generation, review, debugging and repository-scale tasks, especially when paired with a retrieval layer. At $4.57 per million input tokens, it sits in the mid-range of the index, and the 128,000-token context window is enough for large files and multi-file diffs.