Qwen3.7 Flash on QuickSilver Pro
Qwen 3.7 Flash is Alibaba's fast multimodal agent model for visual coding, search, computer interaction, and high-volume automation. It supports text and image input over a 1M-token context. QuickSilver Pro serves it at $0.024 input / $0.104 output per million tokens — 20% below OpenRouter.
At a glance
Fast multimodal agents, visual coding, search, and high-volume automation over a 1M context.
Pricing comparison ($/1M tokens)
| Provider | Input | Output | vs QSP |
|---|---|---|---|
| QuickSilver Pro | $0.024 | $0.104 | lowest-cost |
| OpenRouter (qwen/qwen3.7-flash) | $0.03 | $0.13 | 20% lower |
| OpenAI (GPT-4o mini) | $0.15 | $0.60 | 83% lower |
When to use
Use Qwen 3.7 Flash for high-volume multimodal chat, visual coding, screenshot and document understanding, search agents, tool use, and computer-interaction workflows where latency and price matter. Thinking is suppressed by default for predictable output; opt in when a task needs deeper reasoning.
When to use something else
For the hardest reasoning, step up to Qwen 3.7 Max or DeepSeek V4 Pro. For long-running coding loops where quality matters more than the lowest possible token price, compare Qwen 3.7 Plus.
Quickstart (curl)
curl https://api.quicksilverpro.io/v1/chat/completions \
-H "Authorization: Bearer $QSP_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "qwen3.7-flash",
"messages": [{"role": "user", "content": "Hello!"}]
}'OpenAI-compatible. One-line migration via base_url.
FAQ
Yes. QuickSilver Pro defaults to `reasoning: { enabled: false }` for fast direct answers. Pass `reasoning: { enabled: true }` to receive a reasoning trace and budget additional output tokens.
The current route supports text and image input, text output, tool calling, and a 1M-token context. It is designed for multimodal agents, visual coding, search, and computer interaction.
OpenRouter lists Qwen 3.7 Flash at $0.03 input / $0.13 output per 1M tokens. QuickSilver Pro is $0.024 / $0.104 — 20% below on both legs, with the model ID shortened to `qwen3.7-flash`.