DashScope Aliyun vs SiliconFlow: Qwen API, Free Tier, and Model List

Quick Answer: What Is the Difference Between DashScope and Qwen?

Qwen is Alibaba's model family. DashScope (Aliyun Bailian / Model Studio) is Alibaba Cloud's API platform for calling Qwen and related models. SiliconFlow is a separate inference provider that also serves Qwen variants alongside other Chinese model families. If you only need the official Qwen API path, start with DashScope; if you want a broader model catalog or free prototyping options, compare SiliconFlow too.

Quick Summary: DashScope Model List vs SiliconFlow Breadth

The shortest comparison: DashScope is the official Qwen API and Alibaba Cloud billing path; SiliconFlow is the broader model list for teams that want Qwen plus DeepSeek, GLM, Kimi, OCR, and reasoning models without separate provider accounts.

SiliconFlow integration is rolling out on TheRouter — the SiliconFlow entries below use SiliconFlow's published catalog and pricing rather than current TheRouter routing. DashScope Aliyun is available via TheRouter today.

SiliconFlow — Best for Breadth and Free-Tier Prototyping

SiliconFlow is a high-performance inference platform hosting 200+ models across multiple providers — not just Qwen. SiliconFlow integration is rolling out on TheRouter; the entries below reflect SiliconFlow's own published catalog and pricing. In SiliconFlow's catalog, three models are listed as completely free with no token limits per day:

• Qwen3-8B — fast, compact, supports thinking mode. Good default for everyday tasks.
• DeepSeek-R1-Distill-Qwen-7B — free reasoning model with chain-of-thought output.
• DeepSeek-OCR — free document parsing and OCR model.

Beyond the free tier, SiliconFlow's own catalog also lists DeepSeek V4, Kimi, GLM (Zhipu), and additional Qwen variants through SiliconFlow's OpenAI-compatible endpoint. TheRouter's SiliconFlow integration is rolling out — check the SiliconFlow provider page for what is currently routable.

Best for: teams evaluating broad model access or a zero-cost starting point directly with SiliconFlow, while TheRouter's SiliconFlow integration rolls out.

DashScope Aliyun / Bailian — Official Qwen API and Model List

DashScope Aliyun (阿里云百炼 / Model Studio) is Alibaba Cloud's own API platform — the official source for Qwen models. In practice, developers use DashScope to access the Qwen model list, free-trial quotas, compatible-mode endpoints, and production billing controls. Through TheRouter, it's accessible via both the China endpoint ( dashscope.aliyuncs.com, Beijing) and the international endpoint (dashscope-intl.aliyuncs.com, Singapore), with automatic cheapest-region routing on every request.

DashScope model list highlights available through TheRouter:

• qwen3-max — top-tier Qwen3 model for complex reasoning and coding.
• qwen3-coder-plus / qwen3-coder-flash — Qwen code-specialized variants for agentic coding and IDE workloads.
• QwQ-32B — Alibaba's dedicated reasoning / thinking model.
• qwen-long — long-context model for document-heavy workloads.
• Wan and Qwen image models — text-to-image and image-editing families for media workflows.
• text-embedding-v3 / v4 — two generations of Qwen embeddings for RAG pipelines.

Best for: teams that need the full commercial Qwen3 lineup, CN-cost pricing for China-hosted workloads, Wan image generation, authoritative Qwen embeddings, or a clearer path from the DashScope free tier into production usage.

What Is the Free Tier of DashScope?

The honest answer to “what is the free tier of DashScope?” is: treat it as a trial quota, not a durable production plan. Alibaba Cloud commonly grants new Model Studio users per-model trial tokens with an expiry window, and some open-source Qwen models may appear as free or promotional in the console. The exact quota changes by account type, region, and model family, so verify the current dashscope.aliyun.com / Aliyun Bailian console before promising free usage to your own users. Use the DashScope free tier for smoke tests, SDK migration, and model evaluation; use TheRouter's ledger when you need stable team-level spend visibility across DashScope and non-DashScope models.

OpenAI Compatibility: dashscope.aliyun.com Without SDK Rewrites

Native DashScope compatible mode usually starts from Alibaba's dashscope.aliyun.com endpoint. On TheRouter, DashScope Aliyun uses one OpenAI-compatible base URL, so you only change the model name. SiliconFlow's own API is also OpenAI-compatible — call it directly until TheRouter's SiliconFlow integration is live:

curl https://api.siliconflow.cn/v1/chat/completions \
  -H "Authorization: Bearer $SILICONFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Qwen/Qwen3-8B",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

curl https://api.therouter.ai/v1/chat/completions \
  -H "Authorization: Bearer $THE_ROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen/qwen3.8-max",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

The request body format is the same OpenAI-compatible shape for both. Stream, function calling, and JSON mode are supported on DashScope via TheRouter today; check SiliconFlow's own docs for parity until the TheRouter integration lands.

When to Use Each Qwen API Provider

Routing DashScope Today, SiliconFlow as It Rolls Out

DashScope Aliyun is available on TheRouter today with one API key and unified billing. SiliconFlow's TheRouter integration is rolling out — once it lands, you'll be able to combine SiliconFlow's free Qwen3-8B for lightweight tasks with DashScope's QwQ-32B or qwen3-max for complex reasoning under the same key and billing. Until then, call SiliconFlow directly and route DashScope through TheRouter.

TheRouter adds automatic failover for its integrated providers: if one returns a 5xx or quota error, the router retries on an equivalent model without any changes to your code.