DashScope Qwen API Pricing: 6 Models Compared by Real Token Cost
·6 min read
Answer first: the lowest tracked DashScope input rate in this 2026-07-14 snapshot is Qwen3 at about $0.0138 per million tokens, while Qwen3 Max is the highest-priced record at about $0.3439 input and $1.3755 output. These are maintained ComputeUnion records rather than live API quotes, so verify the source and update time on the platform page before purchasing.
DashScope Qwen pricing snapshot
| Model | Input / 1M | Output / 1M | 10M input + 2M output |
|---|---|---|---|
| Qwen3 | $0.0138 | $0.0550 | $0.25 |
| Qwen3 Max | $0.3439 | $1.3755 | $6.19 |
| Qwen Long | $0.0688 | $0.2751 | $1.24 |
| Qwen Plus | $0.1100 | $0.2751 | $1.65 |
| Qwen Turbo | $0.0413 | $0.0825 | $0.58 |
| QwQ | $0.1100 | $0.3301 | $1.76 |
Monthly scenario = 10 million input tokens plus 2 million output tokens. Rounded totals use the tracked rates above. They exclude caching, retries, tools, taxes, and provider markups.
What the numbers mean
Qwen3 is the lowest-cost record in this snapshot. Qwen Turbo is the next low-cost option. Qwen3 Max costs substantially more, so it should be selected because its results fit the workload—not merely because its name indicates a higher tier.
The official API vs relay pricing decision should include more than the displayed token rate. Compare source status, billing currency, rate limits, data handling, support, availability, and the latest update time. A channel quote can be useful, but it is not automatically better than the official route.
Which Qwen API should you test?
- Cost-sensitive batch jobs: start by testing Qwen3 or Qwen Turbo.
- Long-input workloads: include Qwen Long in the evaluation, then measure actual token use.
- General production work: compare Qwen Plus with the lower-cost options on your own prompts.
- Higher-tier workloads: test Qwen3 Max only when measured output quality can justify the larger bill.
- Reasoning-oriented evaluation: compare QwQ against current Qwen alternatives using the same test set.
Clear conclusion
Do not choose a Qwen endpoint from its name alone. Use the DashScope platform pricing page as the current source record, calculate your monthly token mix, and validate at least two models on representative prompts before committing production traffic.
Frequently Asked Questions
What is the cheapest tracked Qwen API on DashScope?
In the 2026-07-14 ComputeUnion snapshot, Qwen3 has the lowest tracked input rate at about $0.0138 per million tokens. Check the live platform page before purchasing.
How much does Qwen3 Max cost in this snapshot?
The tracked rate is about $0.3439 per million input tokens and $1.3755 per million output tokens.
Are these live DashScope API prices?
No. The platform records are currently maintained rather than live_api sourced. Always verify source status and update time.
Should I choose an official API or a relay?
Compare price source, limits, data handling, support, availability, and update time. The cheapest relay quote is not automatically the safest production choice.