Model catalog

Grok 4.20 Non-reasoning

API

Grok 4.20 Non-reasoning is a language model from SpaceXAI with a 1M-token context window. Provider list price is $1.25 per million input tokens and $2.50 per million output; on Allocate you pay $1.34 and $2.67. It is a closed model served over API; the weights are not published.

Pricing

Provider listOn Allocate
Input, per M tokens$1.25$1.34
Output, per M tokens$2.50$2.67
Cached input, per M tokens$0.20$0.21

Prices checked 2026-07-21.

Price against its peers

Kimi K2.5 Fp4$0.50$2.80
Grok 4.20 Non-reasoning$1.25$2.50
Grok 4.3$1.25$2.50
InputOutput

Provider list prices per M tokens, Grok 4.20 Non-reasoning against its nearest language peers by price.

What a real workload costs

Take 1,000,000 requests a month at 1,200 input and 350 output tokens each: 1,200M input and 350M output tokens. At list prices that is 1,200 × $1.25 + 350 × $2.50 = $2,375 a month. Billed on Allocate it is $2,541.

Grok 4.20 Non-reasoning is served over API. Route traffic to it by name, meter every token, and swap it out in one click when a better fit ships.

Example usage

Point a route at xai/grok-4.20-non-reasoning and the endpoint never changes; swap the model behind it whenever you want.

api.allocate.network
curl https://api.allocate.network/v1/chat/completions \
  -H "Authorization: Bearer $ALLOCATE_KEY" \
  -d '{
    "model": "xai/grok-4.20-non-reasoning",
    "messages": [{"role": "user",
      "content": "Summarise the attached contract."}]
  }'
200 · xai/grok-4.20-non-reasoning · inside your boundary

Common questions

How much does Grok 4.20 Non-reasoning cost per million tokens?

Provider list price is $1.25 per million input tokens and $2.50 per million output tokens. On Allocate you pay $1.34 in and $2.67 out.

What context window does Grok 4.20 Non-reasoning have?

1,000,000 tokens (1M). At roughly 0.75 words per token, that is about 750k words of English text per request.

What does cached input cost on Grok 4.20 Non-reasoning?

$0.20 per million tokens at list ($0.21 billed). Repeated prompt prefixes, such as a stable system prompt or tool definitions, bill at this rate instead of the full input price.

Can I fine-tune Grok 4.20 Non-reasoning?

No. Grok 4.20 Non-reasoning is a closed model served over API; the weights are not published. If you want a model you can train and own, start from an open-weights base in the catalog and fine-tune that.

How do I call Grok 4.20 Non-reasoning on Allocate?

Send xai/grok-4.20-non-reasoning in the model field of the OpenAI-compatible endpoint at api.allocate.network/v1, or point a route name (like prod/support-agent) at it so you can swap the model later without a deploy.

Compare against