Every current Grok model with its real per-million-token price — plus the long-context surcharge, a calculator for your own workload, and a straight comparison against the frontier rivals xAI is undercutting.
| MODEL | INPUT $/1M | CACHED IN | OUTPUT $/1M | BLENDED | CONTEXT | YOUR COST |
|---|---|---|---|---|---|---|
Grok 4.7 grok-4.7 · ≥200K: $4/$12 | $2.00 | $0.50 | $6.00 | $3.00 | 500K | $5.00 |
Grok 4.6 grok-4.6 · ≥200K: $4/$12 | $2.00 | $0.50 | $6.00 | $3.00 | 500K | $5.00 |
Grok 4.3 grok-4.3 · ≥200K: $2.50/$5 | $1.25 | $0.20 | $2.50 | $1.56 | 1M | $2.50 |
Grok Build 0.1 grok-build-0.1 · coding | $1.00 | $0.20 | $2.00 | $1.25 | 256K | $2.00 |
USD per 1M tokens from xAI’s official pricing page as of October 16, 2026. Blended assumes a 3:1 input:output ratio. Get keys at the xAI console.
Grok 4.7, released in September, keeps xAI’s flagship sticker at $2 input / $6 output per 1M tokens — the lowest headline rate of any closed-lab flagship, level with GPT-6 Sol on input, 40% below it on output, and well under Claude Opus 5.5. The previous flagship 4.6 stays at the same price, and Grok 4.3 hangs around as the value pick: $1.25 / $2.50 with a full 1M-token window.
The catch mirrors Gemini: a long-context surcharge. Once a prompt reaches 200K tokens, every Grok text model bills the whole request at double rates — $4 / $12 on 4.7 and 4.6, $2.50 / $5 on 4.3, $2 / $4 on Grok Build — cached input included. Cached input helps on the other side — repeated prefixes bill at about a quarter of the fresh rate ($0.50 on 4.7).
Grok Build 0.1 is the coding-tuned model at $1 / $2 with 256K context — priced to compete with open-weight coding models like Kimi K2.7 Code rather than with the closed-lab rivals.
Grok bills per token with separate input and output rates per million tokens: Grok 4.7 and 4.6 at $2 / $6, Grok 4.3 at $1.25 / $2.50, and the coding-focused Grok Build 0.1 at $1 / $2. Prompts of 200K tokens or more bill at double rates on every model.
Grok Build 0.1 at $1 input / $2 output per 1M tokens — but it is tuned for coding. For general work, Grok 4.3 at $1.25 / $2.50 with a 1M-token context is the value pick.
Requests whose prompts reach 200K tokens bill at double rates: $4 / $12 on Grok 4.7 and 4.6 (up from $2 / $6), $2.50 / $5 on Grok 4.3, and $2 / $4 on Grok Build. Budget at the surcharge rate if you regularly send very large contexts.
At sticker price, yes: Grok 4.7 blends to about $3/1M tokens versus $4 for GPT-6 Sol and $8 for Claude Opus 5.5. Output-heavy workloads benefit most, since Grok charges $6 per 1M output tokens against $10–20 for the rivals. The matchup table below puts them side by side.
From the xAI console at console.x.ai — usage is pay-as-you-go per token. Grok models are also available through OpenRouter if you want one key across many vendors.
Also see: OpenAI API pricing · Claude API pricing · Gemini API pricing · DeepSeek API pricing · Kimi API pricing · GLM API pricing · MiniMax API pricing · MiMo API pricing · all 68 models
ReqKey gives every key a credit balance — validate, deduct, enforce limits, and watch per-consumer Grok spend in real time.