xAI Grok 4.7 API
Grok 4.7 是 xAI 面向编程、Agent 与知识工作的旗舰模型，采用更大的基础模型与更长强化学习训练，强化长任务规划、自我验证、长上下文管理以及文档与演示创作。支持文本与图像输入、文本输出、500K Token 上下文、low/medium/high/xhigh 四档推理（默认 high）、函数调用、结构化输出、流式响应、提示缓存，以及 Web 搜索、X 搜索和代码执行；知识截止时间为 2026 年 5 月。官方评测包括 CursorBench 4.0 46.3%、DeepSWE v1.1 71.0%（high）、EEBench 64.0% 和 Terminal-Bench 4.0 38.0%。
Starting price 按实际 Token 用量计费 · Price snapshot
Waiting to refresh live prices…
Model pricing
Compare all public tiers and current prices for this series on one page.
Input
按实际 Token 用量计费
Output
总价 = 各类 Token 用量 × 对应单价
Context
500,000
Max output
—
- Model ID
-
grok-4.7 - Billing model
- Fixed combination billing
- Available groups
- default, svip
- Capabilities
- chat, responses-native, structured-outputs
- Combination pricing rules
按实际 Token 用量计费总价 = 各类 Token 用量 × 对应单价
- Token 单价由当前计费表达式推导；实际费用按请求用量结算
查看高级计费表达式
v2:len < 200000 ? tier("xai_grok_4_7_short_official_usd_2026_09_22", tokens(p, 2) + tokens(c, 6) + tokens(cr, 0.5) + calls(tool("web_search"), 5) + calls(tool("code_execution"), 5)) : tier("xai_grok_4_7_long_ge_200k_official_usd_2026_09_22", tokens(p, 4) + tokens(c, 12) + tokens(cr, 1) + calls(tool("web_search"), 5) + calls(tool("code_execution"), 5))
API reference
Base URL is https://api.rokoapi.com/v1. The model in requests must use the model IDs listed above.
curl https://api.rokoapi.com/v1/chat/completions \
-H "Authorization: Bearer $API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "grok-4.7",
"messages": [{"role":"user","content":"Introduce yourself in one sentence"}]
}'For more parameters, limits, and async task queries, read the xAI Grok 4.7 developer guide. xAI Grok 4.7 Developer guide · OpenAI · Authorization