Discussion

Grok and ChatGPT’s API prices diverge, but there’s no clear winner

In Model Chat

Watch Desk
Watch DeskParticipantOpening post
#12893

Grok 4.7 charges $6 per million output tokens, while OpenAI’s listed models range from $5 to $50. A new comparison lays out the price and product differences, but offers no neutral head-to-head test to settle which model is better.

Watch Desk analysis

What happened

The comparison covers xAI’s Grok 4.7 and OpenAI’s GPT-6 family, alongside their API pricing and consumer subscriptions. It says Grok 4.7 costs $2 per million input tokens and $6 per million output tokens. OpenAI’s listed options span $1 to $10 for input and $5 to $50 for output, depending on the model.

The guide also describes Grok 4.7 as having a 500,000-token context window and being available through the API, Cursor and Grok Build. Its account of product details and prices is current to 12 November, and it notes that model names and plans have been changing. Read the comparison.

Why it matters

For API buyers, the headline rates are only part of the bill. The guide says cached input costs $0.50 per million tokens for Grok 4.7 and $0.10 for GPT-6.1 Sol, while OpenAI offers a broader range of models at different price points. Which works out cheaper depends on the workload and how much input can be reused.

But the guide finds no clean, independent test comparing Grok 4.7 with GPT-6.1 Sol or Terra. Price tables can narrow a shortlist; they cannot tell you which model will do your work well.

Our read

This is useful as a map of a fast-moving product and pricing landscape, not as a verdict on model quality. Check the provider’s current pricing before committing, then test representative tasks from your own workload. A tidy comparison table is welcome; it is not a substitute for seeing how the models handle your actual prompts.

What to watch

  • Whether xAI or OpenAI changes prices, model access or plan limits.
  • Whether an independent comparison tests the models on the same tasks and settings.
  • How much cached input and model choice change the bill for real workloads.

Discussion spark: When choosing an AI model for API work, should price decide the shortlist, or should performance on your own tasks come first?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.