AI modelsxAI

Grok 4.3

Decision matrix

Grok 4.3 model decision matrix
Overview

Grok is xAI's model family, called through its API with pricing that scales by tier and context length; the flagship grok-4.7 runs $2/$6 per million input/output tokens (500K context), with cheaper tiers like grok-4.3 at $1.25/$2.50 (1M context). Prices double for prompts over 200K tokens. Pick the tier that matches your task's context needs, not just its raw capability.

Other good fits
  • A task that fits comfortably under 200K tokens, avoiding the extended-context pricing tier that doubles the rate.
Pros
Read pros

Grok is xAI's model family, accessible through its own API with several tiers trading capability, context window and price against each other. Unlike a single flagship price, Grok's tiers are genuinely differentiated: a mid-tier model with a larger context window can cost less per token than the flagship with a smaller one, so picking the right tier for the task matters more here than with some other providers.

Pick it when…

  • You want xAI's models specifically as an alternative or complement to OpenAI, Anthropic or Google's offerings.
  • Your task's context needs and price sensitivity point toward a specific tier rather than defaulting to the most expensive one.
  • You're building something where image or video generation (also part of xAI's API) might sit alongside text generation in the same account.
Cons
Read cons

The extended-context pricing doubling above 200K tokens can catch you by surprise if you're not tracking prompt length carefully, especially on a task that grows over time (a long conversation history, for instance). Tier naming and pricing here change relatively often; confirm current tier names and rates before committing a production budget.

Pick something else when…

  • You want the most stable, long-established API ecosystem. Claude API or OpenAI's API have more mature third-party tooling.
  • You want to compare Grok against several other providers from one integration. OpenRouter makes that easy.
  • Your prompts regularly exceed 200K tokens and cost predictability matters. Check whether a competing provider's flat-rate pricing works out cheaper for your specific usage pattern.

Avoid if

  • Your prompts regularly exceed 200K tokens; pricing doubles above that threshold on every model.
Context

Not documented yet.

Requirements

Not documented yet.

Licence & cost

Prices checked 23 September 2026 (API pricing (opens in a new tab)), per million tokens, under 200K-token prompts:

ModelInputOutputContext
grok-4.7 (flagship)$2.00$6.00500K
grok-4.3$1.25$2.501M
grok-build-0.1$1.00$2.00256K

Prices double for the portion of a prompt at or beyond 200K tokens, on every tier. xAI's API also covers image generation (from $0.02/image) and video generation (from $0.05/second) if your project needs those alongside text. There's no subscription for API access; billing is pay-as-you-go per token or per generation.

First task

Documentation-based suggestion — not run here.

Send the same made-up drafting task to grok-4.7 (the flagship) and grok-4.3 (a cheaper tier with a larger context window), and compare quality against the roughly 40% lower price of the cheaper tier.

Done when you have a real side-by-side comparison and can say which tier you'd default to for similar future tasks, rather than assuming the most expensive option is automatically the right choice.

Links & evidence
Documentation-based · not hands-on · Sep 23, 2026

Rests on xAI's API models and pricing documentation, read on 2026-09-23. No API call was made for this review; tier names and prices change relatively often for this provider, so confirm current rates before budgeting.

  • Grok API models and pricing · Sep 23, 2026 — grok-4.7 $2/$6 (500K context), grok-4.3 $1.25/$2.50 (1M context), extended-context doubling above 200K tokens.
  • OpenRouter pricing · Sep 23, 2026 — Comparison alternative for calling Grok alongside other vendors.
  • Claude API pricing · Sep 23, 2026 — Comparison alternative with more established ecosystem tooling.
Further review notes & alternatives

Connection

Call Grok models through xAI's API with your API key, specifying the model name (for example grok-4.7 or grok-4.3). Pricing is tiered by prompt length: the quoted rate applies under 200K tokens, and doubles for prompts at or above that threshold, so a single long-context request can cost significantly more than the headline rate suggests.

Alternatives and trade-offs

  • Claude Sonnet wins on ecosystem maturity and documentation depth as a cross-vendor comparison.
  • GPT-5 wins as another direct flagship comparison; worth benchmarking all three on your actual task.
  • OpenRouter wins when you want to try Grok alongside several other providers without separate integrations.

Learn more

Catalogue facts

Analyze text and video with a large context window. Use live X data when current social posts are relevant evidence. Choose it when cloud access and X integration fit the task.

xAI · paid · cloud

  • frontier model
  • video input
  • live twitter data
  • large token context
  • low per-token cost
  • always-on reasoning
  • high-volume tier
  • talk-think
  • search-cite
  • write-code
  • draft-docs

Browse all tools