Provider pricing hub

Grok API pricing: xAI Grok 4.7, 4.6 and 4.5.

Grok API prices on RouterPlex run from $2.00 / $6.00 per 1M tokens, with 0% markup, below 200K prompt tokens. Longer prompts bill at $4.00 / $12.00. No xAI account needed. Top-ups have a $0 fee, by card from $5 or crypto; balances stop at $0 and each key has a hard USD cap.

Prices checked 17 November 2026 from the live catalog. xAI's pricing page last checked 10 Nov 2026: docs.x.ai/docs/models

Grok on this key17 November 2026

4

Routes

$2.00

Cheapest input / 1M

Markup 0%

xAI Grok family

Published catalog rates.

Agent-turn cost uses 100K input and 10K output, below the long-context threshold.

Model routeContextInput / 1MCached input / 1MOutput / 1MAgent turn
Grok 4.5

grok-4.5

500K$2.00not offered$6.00$0.26
Base rate applies below 200K prompt tokens. Prompts at or above 200K are billed at $4 input / $12 output per 1M tokens. Cached prompt tokens are billed at the full input rate in both tiers: our upstream charges more for a cached grok token than a fresh one.
Grok 4.6

grok-4.6

500K$2.00not offered$6.00$0.26
Base rate applies below 200K prompt tokens. Prompts at or above 200K are billed at $4 input / $12 output per 1M tokens. Cached prompt tokens are billed at the full input rate in both tiers ($2, or $4 at or above 200K): our upstream charges more for a cached grok token than a fresh one, so there is no cache discount to pass through on this model.
Grok 4.7

grok-4.7

500K$2.00not offered$6.00$0.26
xAI list price. Base rate applies below 200K prompt tokens; prompts at or above 200K are billed at $4 input / $12 output per 1M tokens. Cached prompt tokens are billed at the full input rate for now: RouterPlex does not yet pass xAI's $0.50 cache-read tier through on this model.
Grok 4.7 Build Fast

grok-4.7-build-fast

500K$4.00not offered$12.00$0.52
Grok 4.7 on xAI's faster serving tier, which xAI sells inside Grok Build rather than on its public API. RouterPlex bills xAI's Fast rate, 2x Grok 4.7: $4 input / $12 output per 1M tokens below 200K prompt tokens and $6 / $18 at or above 200K. Cached prompt tokens are billed at the full input rate for now.

No Grok route in this catalog lists a cached-input rate, so that column reads "not offered".

Grok pricing calculator

Same workload. Every Grok route.

Defaults are a 10K-in / 1K-out request, 3,000 times a month. Set input to 200,000 or more to see the long-context rate the bill switches to.

Workload calculator

Catalog rates · USD · no markup

Models in the estimate

4 selected
10K in + 1K out × 3,000 req$100 of tokens
RoutePer requestMonthly$100 buys
Grok 4.5

xAI

$0.026$78.003,846 req
Grok 4.6

xAI

$0.026$78.003,846 req
Grok 4.7

xAI

$0.026$78.003,846 req
Grok 4.7 Build Fast

xAI

$0.052$156.001,923 req

Estimates apply live catalog rates and switch to a route's long-context rate when the prompt reaches its threshold. Actual spend follows the tokens your client sends and the model returns. Image-generation routes are omitted because they are not a chat turn.

The 200K cliff

What happens above 200K prompt tokens.

xAI bills every token in a request at the higher rate once the prompt reaches the threshold, not only the tokens past it. Checked 2026-11-10 on docs.x.ai/docs/models.

Grok 4.5 requestRate / 1MCost
190K in + 10K out$2.00 / $6.00$0.44
210K in + 10K out$4.00 / $12.00$0.96

20K more prompt tokens cost 2.2x as much. Trim or summarize context before it crosses the line. The calculator above switches to the long-context rate when the input field reaches 200K.

Call it

Use Grok from Claude Code or the OpenAI SDK.

Claude Code and the Anthropic SDK use https://api.routerplex.com. The OpenAI SDK, Cursor and Cline use https://api.routerplex.com/v1. Same key, same catalog id, same per-key budget. Swap grok-4.7 for any id in the table.

Claude Code (shell)
export ANTHROPIC_BASE_URL=https://api.routerplex.com
export ANTHROPIC_AUTH_TOKEN="$ROUTERPLEX_API_KEY"
export ANTHROPIC_MODEL=grok-4.7
claude
OpenAI SDK (Python)
import os
from openai import OpenAI
 
client = OpenAI(
base_url="https://api.routerplex.com/v1",
api_key=os.environ["ROUTERPLEX_API_KEY"],
)
reply = client.chat.completions.create(
model="grok-4.7",
messages=[{"role": "user", "content": "Hello!"}],
)
print(reply.choices[0].message.content)

How to pick

Same rate, three versions: watch the context.

These notes only use catalog fields: price, context, and mode. They are not quality rankings.

Price does not pick the version

Compare the rows in the table; versions are priced differently.

The cliff costs more than the version

Crossing 200K prompt tokens moves the whole request to $4.00 / $12.00. Keeping prompts under it saves more than any version switch.

No xAI account

One RouterPlex key reaches Grok, Claude, GPT, Gemini, DeepSeek, Qwen and GLM. Claude Code uses https://api.routerplex.com; the OpenAI SDK uses https://api.routerplex.com/v1.

Setup notes for Grok 4.7, Grok 4.6 and the longer Grok API pricing write-up. Compare the rest of the catalog on all model pricing, the cheapest-first ledger, or RouterPlex vs OpenRouter.

Before you bill production

Grok pricing, in questions.

01How much does the Grok API cost?+

As of 17 November 2026, Grok on RouterPlex costs $2.00 / $6.00 for Grok 4.5, $2.00 / $6.00 for Grok 4.6, $2.00 / $6.00 for Grok 4.7 and $4.00 / $12.00 for Grok 4.7 Build Fast per 1M input / output tokens with 0% markup. Top-ups have a $0 fee and the balance stops at $0.

02What is the Grok 200K cliff?+

Once a prompt reaches 200,000 tokens, the whole request bills at $4.00 / $12.00 per 1M instead of $2.00 / $6.00. A 190K-in / 10K-out request costs about $0.44; 210K in costs about $0.96. (Source: docs.x.ai/docs/models, checked 2026-11-10.)

03Which Grok model is newest?+

Grok 4.7 Build Fast ("grok-4.7-build-fast") is the highest Grok version in this catalog, with 500K context at $4.00 / $12.00 per 1M.

04How do I get a Grok API key?+

On RouterPlex you do not need a xAI account: sign up, verify your email, and create a key in the dashboard (you can set a hard USD budget on it). Call https://api.routerplex.com/v1 with the OpenAI SDK, or point Claude Code at https://api.routerplex.com, with model "grok-4.7". The same key reaches every other model in the catalog. For a direct xAI key, xAI's quick start says to sign up at console.x.ai, load it with credits, and create a key on the API Keys page. (Source: docs.x.ai/docs/tutorial, checked 2026-11-10.)

05Is there a free Grok API tier on RouterPlex?+

No. RouterPlex has no free tier. Eligible verified accounts get $1 of test credit once, no card, and it never expires. After that you top up from $5 by card, or with crypto, with a $0 top-up fee; the balance stops at $0, so a Grok loop cannot run up a bill.

06Which Grok model id should I use?+

Send the catalog id exactly as written: "grok-4.5" (Grok 4.5), "grok-4.6" (Grok 4.6), "grok-4.7" (Grok 4.7) and "grok-4.7-build-fast" (Grok 4.7 Build Fast). The cheapest per agent turn is "grok-4.5". The page URL slug (for example /models/grok-4-5) is not the API id; the id keeps its dots and case. Anything else returns model_not_found.

07Does Grok have cached-input pricing here?+

Not today. No featured Grok route in this catalog lists a cached-input rate, so repeated prompt tokens bill at the full input rate. xAI publishes a discounted cached-input rate on its own platform, but each Grok row's catalog note says RouterPlex bills cached Grok tokens at the full input rate for now. (Source: docs.x.ai/docs/models, checked 2026-11-10.)

08Can I use Grok with the OpenAI SDK or Claude Code?+

Yes. The OpenAI SDK uses https://api.routerplex.com/v1 with your RouterPlex key and model "grok-4.7". Claude Code uses https://api.routerplex.com with the key as ANTHROPIC_AUTH_TOKEN.

Route Grok from one balance

Try Grok 4.7 with $1, no card.

Try it with $1 of test credit, no card. Guided setup shows the exact Claude Code config for Grok 4.7.

Eligible verified accounts get the $1 once. It never expires.

Grok API Pricing (Nov 2026): xAI Grok 4.5-4.7 · RouterPlex