xAI

Grok 4.6

xAI frontier model for agents, coding, and visual projects

Grok 4.6 is an xAI multimodal frontier model for long-running agents, coding, knowledge work, reasoning, tool use, and visual projects, with a 500K context window. Best for production jobs that need long-horizon execution, tool collaboration, and mixed text/visual input.

500,000-token context500K
Released2026-08-12
Relays78 sites
Agentic tasksCodingKnowledge workReasoningTool use500K context

xAI Official Pricing

CNY
Updated: 2026-08-19T12:34:41.697+08:00Source

observed_hvoyweb-lte-200k · Input

~¥14.4/ 1M tokens

Observed from the hvoyweb active catalog and LLM metadata on 2026-08-19; this tier applies at up to 200K total input, with USD prices converted using the app_config FX rate.

observed_hvoyweb-lte-200k · Output

~¥43.2/ 1M tokens

Observed from the hvoyweb active catalog and LLM metadata on 2026-08-19; this tier applies at up to 200K total input, with USD prices converted using the app_config FX rate.

observed_hvoyweb-lte-200k · Cache read

~¥3.6/ 1M tokens

Observed from the hvoyweb active catalog and LLM metadata on 2026-08-19; this tier applies at up to 200K total input, with USD prices converted using the app_config FX rate.

observed_hvoyweb-gt-200k · Input

~¥28.8/ 1M tokens

Observed from the hvoyweb active catalog on 2026-08-19; this long-context tier applies above 200K total input, with USD prices converted using the app_config FX rate.

observed_hvoyweb-gt-200k · Output

~¥86.4/ 1M tokens

Observed from the hvoyweb active catalog on 2026-08-19; this long-context tier applies above 200K total input, with USD prices converted using the app_config FX rate.

observed_hvoyweb-gt-200k · Cache read

~¥7.2/ 1M tokens

Observed from the hvoyweb active catalog on 2026-08-19; this long-context tier applies above 200K total input, with USD prices converted using the app_config FX rate.

Observed from the hvoyweb active catalog and LLM metadata on 2026-08-19; this tier applies at up to 200K total input, with USD prices converted using the app_config FX rate.

Relay Comparison

Compare token, per-request, or per-second pricing by relay channel.

How should Grok 4.6 relay pricing be compared?

This Grok 4.6 pricing page compares official pricing with public prices from 78 listed AI gateways. Token prices are shown in CNY per 1M tokens, while per-request, per-second, and per-character rows use the unit shown in the table. Last updated: 09/20/2026, 23:14.

Data sources
Public price catalogs, official pricing records, and monitoring results.
Metric definitions
Uptime means successful probe response rate, fake-rate signals possible model mismatch or abnormal output risk, and latency is average API response time.
Risk note
Relay gateways are third-party services. Pricing, billing, privacy, and stability can change; start with a small top-up and verify reliability before continued use.