xAI Lists Grok 4.7: $1.60 Input, 500k Context Window

What's new

xAI quietly listed Grok 4.7 on 2026-09-21, adding another multimodal option to a crowded mid-tier. The model accepts text, image, and file inputs and returns text — a practical combination for document-heavy or vision-assisted workflows.

Pricing and context

At $1.60 per 1M input tokens and $4.80 per 1M output tokens, Grok 4.7 sits in the affordable-but-not-cheap bracket. The 500k-token context window is the headline number here — large enough to ingest long codebases, legal documents, or multi-file uploads in a single pass without chunking gymnastics.

For a typical RAG workload that processes 200k tokens of context and generates 10k tokens of output, you're looking at roughly $0.37 per run — reasonable, though output costs will bite if your use case is generation-heavy.

Who should care

• Long-document analysts and legal/research teams who need to keep entire files in context will find the 500k window genuinely useful.
• Developers building multimodal pipelines (image + file ingestion) gain another vendor option, which matters for redundancy and cost negotiation.
• Cost-sensitive teams should compare against alternatives on our index: at $4.80/M output, heavier generation tasks may favour competitors pricing output closer to $1–2/M.

Check current specs, context limits, and price history at Grok 4.7 — live specs & price history before committing to a workload.

Bottom line

Grok 4.7 is a solid addition if the 500k context window solves a real problem for you. The pricing is competitive on input but watch output costs on generation-heavy tasks.

Ezra, Scout AI Team