Choose by capabilities, context and access conditions; vendor benchmarks do not establish an independent quality rank.
Keep in mind
Free-credit amounts conflict. Region and account eligibility require official confirmation.
Entitlements and usage
Allowance
One-time signup credit. Official amounts conflict; no confirmed free-token quantity is published.
Benefits
Low-latency reasoning, coding, agents and RAG
Models listed in source
Mercury 2.5, Mercury Voice, Mercury Decide
Source notes and limits
Pay for actual token usage. Voice enterprise access, the decisions endpoint and legacy models have separate conditions.
API input and output pricing
Mercury 2.5
Input: 0.04 USD/1,000,000 Tokens; Output: 0.15 USD/1,000,000 Tokens; Cache read: 0.004 USD/1,000,000 Tokens Official limited-time offer; the end follows official announcements. Publicly documented Inception direct API; account and API key required. Purchase and authenticated inference were not tested. Region eligibility requires confirmation. Inception direct API only; the promotion end date has not been announced. Price evidence valid until 2026-10-22T10:02:53.646Z (not the campaign end). Context: 260K tokens (official decimal display) Verified: 2026-10-15T10:02:53.6464958Z
Supported models
These are reference API prices for the models, not additional plan charges or rates for consuming plan allowances. Only models with records in the current catalog are shown. Check the official plan terms for model availability, allowances and limits.
Prices retain their original currency (¥ CNY / $ USD) and are per million tokens; cache storage is per million tokens per hour. Unpublished prices are marked Pending or —.
Inception Labs Inception Mercury API (pay as you go) — Supported models and reference prices
Model
Brand
Context
Input
Cache read
Output
Mercury 2.5 ↗Offer status: active · Ends (exclusive): Unverified · 官方直连 API 文档公开，需注册账户和 API Key；未执行付费或鉴权推理测试，地区资格待确认。
