benchrRouting note: An earlier benchr snapshot recorded changes to legacy Grok aliases. Verify the exact model returned and billed in current xAI logs rather than assuming an alias still resolves the same way. See the dated alias-and-pricing record.
Pricing breakdown
| Tier | Rate / 1M tokens |
|---|---|
| Standard input | $1.25 |
| Standard output | $2.50 |
| Context window | 1,000,000 tokens |
Input and output price ratio
In this snapshot, GPT-5 lists $10/1M output against $1.25/1M input, while Claude Opus 4.8 lists $25/$5 and Grok 4.3 lists $2.50/$1.25. At an identical 5M-token output volume, the output line item is $12.50 for Grok 4.3 versus $50 for GPT-5. That is rate-card arithmetic, not cost per accepted task: model quality, retries, search-tool charges, caching, and different output lengths can change the operational result.
Native real-time web search
xAI documents Web Search and X Search tools for supported API workflows. Enabling a search tool can give the model access to retrieved current material, but it does not guarantee factual accuracy, complete coverage, or correct citations. Evaluate the sources returned, timestamp handling, corroboration, latency, and total tool charges. Search over the public web also does not replace retrieval over private documents or organization-specific access controls.
Cost scenarios
At 10M input + 10M output per month, the listed Grok 4.3 token rates total $12.50 + $25 = $37.50/month. The comparison rates in this snapshot total $112.50 for GPT-5 and $180 for Claude Sonnet 4.6 at the same token counts. This excludes search tools, retries, caching, and differences in task acceptance or generated length.
Use-case fit
Test for: Long-context requests; report or documentation pipelines with high measured output volume; and tasks where xAI's search tools may add useful public-web or X coverage.
Test alternatives when: You need provider-published coding benchmarks, a known maximum output limit, retrieval over private data, or existing platform controls that would make migration costly.
Decision checklist
Calculate your input-to-output token ratio in production. If your workload generates 2+ output tokens per input token, Grok 4.3's cheap output pricing creates meaningful savings. If your workload is input-heavy (long documents, short answers), the output pricing advantage diminishes.
For search-enabled work, compare source coverage, timestamps, citation correctness, latency, and full tool cost. Do not infer that a public search tool replaces retrieval and authorization for your own documents.