Grok 4.5 vs Grok 4.6.
model vs model · prices as of
Grok 4.5 and Grok 4.6 split the 5 rounds this table counts: the two list price rows, the cheapest price we can explain, the standard price and context. The weighting you pick decides the rest.
xAIGrok 4.5
$12/M tokens
xAI list price, nothing below it is accounted for · $1.56 per 1,000 pages
500k context · tool calling · reads images
xAIGrok 4.6
$12/M tokens
xAI list price, nothing below it is accounted for · $1.56 per 1,000 pages
500k context · tool calling · reads images
List price in (lower wins)
$2/M
$2/M
List price out (lower wins)
$6/M
$6/M
Cheapest we can explain (lower wins)
none below list
none below list
Standard price (lower wins)
$12/M
$12/M
Cost per 1,000 pages (lower wins)
$1.56
$1.56
Context
500k tokens
500k tokens
Open weights
no
no
Tool calling
yes
yes
Structured output
yes
yes
Reads images
yes
yes
Reasoning
yes
yes
Released
8 Jul 2026
12 Aug 2026
Providers selling it
35
34
Price we measure against
the maker's own
the maker's own
Trains on your prompts (no is better)
no provider to check
no provider to check
Providers selling both · 29
302.AI$12 / $12
Abacus$12 / $12
AIHubMix$12 / $12
Atlas Cloud$12 / $12
Cloudflare AI Gateway$12 / $12
CrossModel$12 / $12
Eden AI$12 / $12
GitHub Copilot$12 / $12
Grok 4.5 then Grok 4.6, per million tokens · showing the 8 cheapest of 29
Go further
Add a third or fourth model, or switch the weighting, in the tool.
What moved in the last seven days
Frequently asked
- Which is cheaper, Grok 4.5 or Grok 4.6?
- Neither: both run at $12 per million tokens, three parts input to one part output, read on 16 Sept 2026.
- Where is each one cheapest?
- Grok 4.5: nobody goes below the maker's own price with a reason we can point at, so $12 is the number to beat. Grok 4.6: nobody goes below the maker's own price with a reason we can point at, so $12 is the number to beat.
- Can I buy both from one provider?
- Yes. 29 providers sell both: 302.AI, Abacus, AIHubMix, Atlas Cloud, Cloudflare AI Gateway, CrossModel, and more. The cheapest of them for the two together is 302.AI.
- Which holds more context?
- Both hold 500k tokens in one call.
Prices exclude tax and are per million tokens, read 16 Sept 2026 from each provider's own catalog. The standard price blends three parts input to one part output; a page is 600 tokens in and 60 out. Bars are scaled to the larger value in each row.