Gemini 3.1 Pro Preview API 价格与成本估算
Google Gemini 3.1 Pro Preview 的 API 定价为：输入 $2/百万 tokens，输出 $12/百万 tokens。缓存命中的输入按 $0.2/百万 tokens 计费。上下文窗口 1M tokens。Batch API 按标准价的 50% 计费。价格核实于 2026-07-27，以官方定价页为准。
输入 / 百万 tokens
$2
输出 / 百万 tokens
$12
缓存输入 / 百万
$0.2
上下文窗口
1M
三档负载月成本估算
| 负载 | 每月 tokens | 标准价 | 50% 缓存命中 | Batch API |
|---|---|---|---|---|
| 轻量（原型/side project） | 5M in + 1M out | $22.00 | $17.50 | $11.00 |
| 中等（生产小流量） | 50M in + 10M out | $220 | $175 | $110 |
| 重度（规模化生产） | 500M in + 100M out | $2200 | $1750 | $1100 |
注意：Prompts over 200K tokens bill at $4 / $18.
价位相近的替代模型
| 模型 | 输入/百万 | 输出/百万 | 中等负载月成本 |
|---|---|---|---|
| Google Gemini 3.1 Pro Preview | $2 | $12 | $220 |
| Anthropic Claude Sonnet 5 | $2 | $10 | $200 |
| OpenAI GPT-5.6 terra | $2.5 | $15 | $275 |
| xAI Grok 4.5 | $2 | $6 | $160 |
| Google Gemini 3.6 Flash | $1.5 | $7.5 | $150 |
中等负载 = 每月 5000 万输入 + 1000 万输出 tokens，标准价、无缓存。价格核实于 2026-07-27。
按你的用量计算
LLM API Pricing Calculator
Compare current API prices for Claude, GPT, Gemini, DeepSeek, Kimi, Grok and other LLMs, and estimate your monthly bill from token volume. Includes prompt-caching prices and batch discounts that most comparison tables miss. Prices last verified 2026-07-27. Runs entirely in your browser — no upload, no signup.
Share of input tokens served from prompt cache
| Model | Input $/M | Output $/M | Cached $/M | Context | Est. monthly |
|---|---|---|---|---|---|
DeepSeek V4 Flash DeepSeek · open-weight | $0.14 | $0.28 | $0.0028 | 1M | $9.80 |
Gemini 3.1 Flash-Lite | $0.25 | $1.5 | $0.025 | 1M | $27.50 |
DeepSeek V4 Pro DeepSeek · open-weight | $0.435 | $0.87 | $0.003625 | 1M | $30.45 |
GPT-5 mini OpenAI | $0.25 | $2 | $0.025 | 400K | $32.50 |
GLM-5 Zhipu (z.ai) · open-weight | $1 | $3.2 | $0.2 | 200K | $82.00 |
Grok 4.20 xAI | $1.25 | $2.5 | $0.2 | 1M | $87.50 |
Kimi K2.7 Code Moonshot · open-weight | $0.95 | $4 | $0.19 | 262K | $87.50 |
Claude Haiku 4.5 Anthropic | $1 | $5 | $0.1 | 200K | $100 |
GPT-5.6 luna OpenAI | $1 | $6 | $0.1 | 1.05M | $110 |
GLM-5.2 Zhipu (z.ai) · open-weight | $1.4 | $4.4 | $0.26 | 200K | $114 |
Gemini 3.6 Flash | $1.5 | $7.5 | $0.15 | 1M | $150 |
Grok 4.5 xAI | $2 | $6 | $0.3 | 500K | $160 |
Claude Sonnet 5 Anthropic | $2 | $10 | $0.2 | 1M | $200 |
Gemini 3.1 Pro Preview | $2 | $12 | $0.2 | 1M | $220 |
GPT-5.6 terra OpenAI | $2.5 | $15 | $0.25 | 1.05M | $275 |
Kimi K3 Moonshot · open-weight | $3 | $15 | $0.3 | 1.048576M | $300 |
Claude Opus 5 Anthropic | $5 | $25 | $0.5 | 1M | $500 |
GPT-5.6 sol OpenAI | $5 | $30 | $0.5 | 1.05M | $550 |
Claude Fable 5 Anthropic | $10 | $50 | $1 | 1M | $1000 |
Prices are per 1 million tokens in USD, standard (non-batch) tier, verified 2026-07-27 against official pricing pages. Claude Sonnet 5: Intro price through Aug 31, 2026; $3 / $15 after. GPT-5.6 sol: Requests beyond the long-context threshold bill at 2× input / 1.5× output. Gemini 3.1 Pro Preview: Prompts over 200K tokens bill at $4 / $18. Grok 4.5: Requests over 200K tokens bill at 2×. Kimi K3: Always-on reasoning; output includes thinking tokens.
Batch discount is applied only to models whose provider offers an async batch tier. Cached-input pricing models the read price; cache-write surcharges (Anthropic) are not included. Long-context surcharges apply above provider thresholds and are noted per model.
FAQ
Gemini 3.1 Pro Preview 的 API 价格是多少？
输入 $2/百万 tokens，输出 $12/百万 tokens，缓存命中输入 $0.2/百万 tokens。注意：Prompts over 200K tokens bill at $4 / $18.（核实于 2026-07-27）
用 Gemini 3.1 Pro Preview 跑一个月大概花多少钱？
以每月 5000 万输入 + 1000 万输出 tokens 的中等负载估算约 $220/月；若 50% 输入命中缓存约 $175/月；离线任务走 Batch API 还可按 50% 计费。
Gemini 3.1 Pro Preview 的缓存计费怎么算？
命中提示缓存的输入 tokens 按 $0.2/百万计费，约为标准输入价的 10%。系统提示词、few-shot 示例等重复前缀是主要受益场景。
和 Claude Sonnet 5 相比谁更便宜？
同样中等负载下，Gemini 3.1 Pro Preview 约 $220/月，Anthropic Claude Sonnet 5 约 $200/月。实际选型还应考虑质量、时延和上下文窗口（Gemini 3.1 Pro Preview 为 1M，Claude Sonnet 5 为 1M）。