Gemini 3.7 Flash on QuickSilver Pro
Gemini 3.7 Flash is Google's newest Flash model for production agents, coding, multimodal analysis, and long-context work. It accepts text and images across a 1M-token context and supports reasoning, streaming, tools, and structured output. Through December 31, 2026, QuickSilver Pro charges $0.6375 input / $3.1875 output per million tokens — 15% below Google's promotional API price.
At a glance
Fast multimodal agents and coding with 1M context, at 15% below Google's 2026 promotional API price.
Pricing comparison ($/1M tokens)
| Provider | Input | Output | vs QSP |
|---|---|---|---|
| QuickSilver Pro | $0.6375 | $3.1875 | lowest-cost |
| Google promotional price (Gemini API promotional rate) | $0.75 | $3.75 | 15% lower |
| OpenAI (GPT-4o) | $2.50 | $10.00 | 68% lower |
When to use
Use Gemini 3.7 Flash for production agents, coding assistants, image understanding, document analysis, tool-calling workflows, and long prompts that benefit from a 1M-token context. It supports both Chat Completions and Responses through the same OpenAI-compatible QSP endpoint.
When to use something else
For simple classification or extraction where cost matters more than reasoning quality, Gemini 3.5 Flash-Lite is cheaper. For text-only premium reasoning at a lower token price, evaluate DeepSeek V4 Pro. The current Gemini 3.7 Flash price is promotional through December 31, 2026, so budget-sensitive production workloads should review pricing before 2027.
Quickstart (curl)
curl https://api.quicksilverpro.io/v1/chat/completions \
-H "Authorization: Bearer $QSP_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gemini-3.7-flash",
"messages": [{"role": "user", "content": "Hello!"}]
}'OpenAI-compatible. One-line migration via base_url.
FAQ
Through December 31, 2026, it costs $0.6375 per million input tokens and $3.1875 per million output tokens. Cached input is $0.06375 per million tokens. These rates are 15% below Google's promotional API prices; pricing will be reviewed before the promotion ends.
Yes. It accepts text and image input and supports streaming, reasoning, function calling, and JSON Schema structured output. Its maximum context is 1M tokens and its maximum output is 65,536 tokens.
Set base_url=https://api.quicksilverpro.io/v1, use your QSP API key, and set model="gemini-3.7-flash". Both `/v1/chat/completions` and `/v1/responses` are supported, including streaming and tool calls.
Yes. QuickSilver Pro continuously monitors the model and verified non-streaming Chat Completions, streaming Chat Completions, the Responses API, streaming Responses, and forced function calling against the production endpoint at launch.