AllocateGrok 4.20 Reasoning vs Grok 4.3
On provider list prices, Grok 4.20 Reasoning costs $1.25 per million input tokens against $1.25 for Grok 4.3: effectively level. Output is $2.50 against $2.50.
Grok 4.20 Reasoning Grok 4.3Specifications and provider list prices from the Allocate catalog, checked 2026-07-21.
What the numbers say
Take 1,000,000 requests a month at 1,200 input and 350 output tokens each. That workload costs $2,375 a month on Grok 4.20 Reasoning and $2,375 on Grok 4.3 at list: a gap of $0.
Common questions
Which is cheaper, Grok 4.20 Reasoning or Grok 4.3?
Grok 4.20 Reasoning, on this workload shape. At list prices it is $1.25/$2.50 per million tokens in and out against $1.25/$2.50 for Grok 4.3. Billed on Allocate: $1.34/$2.67 against $1.34/$2.67.
Which has the bigger context window?
They match: both read 1,000,000 tokens (1M) per request.
Can I fine-tune Grok 4.20 Reasoning or Grok 4.3?
No. Both are closed models served over API. If you want a model you can train and own, start from an open-weights base in the catalog.
Related comparisons
Run the numbers on your workload
Or don’t choose. On Allocate a route name is the contract: point yours at one model today, swap to the other tomorrow, and compare them on your live traffic with per-token metering.