Skip to main content

Grok 4.20 0309 v2 (Non-reasoning)

xAIGrok

Description

Grok 4.20 Beta Non-Reasoning is xAI's non-reasoning variant of the Grok 4.20 model family. Released as a beta on March 9, 2026, it provides fast, direct responses without thinking tokens, featuring a 2 million token context window while optimizing for speed and lower latency.

Release Date
2026-04-07
Parameters
Context Length
Modalities

Capability Radar

22
general
33
coding
78
reasoning
51
scienceest.
30
agents
80
multimodal

Science uses a reasoning proxy when dedicated science benchmarks are unavailable.

Rankings

Domain#RankScoreSource
Code Ranking333
25.0
AA
General Ranking227
46.0
AA
Science117
59.0
AA

Benchmark Scores (LLM Stats)

No benchmark data available

AA Evaluation Indices

Intelligence Index
21.8
Gpqa
0.8
Tau2
0.6
Ifbench
0.5
Scicode
0.3
Hle
0.2
Lcr
0.2
Terminalbench Hard
0.2

LLM Stats Category Scores

Long Context
10

Pricing

Input Price$2 / 1M tokens
Output Price$6 / 1M tokens
Blended Price (3:1)$3 / 1M tokens

Speed

Tokens/sec225.6
Time to First Token0.40s
Time to Answer0.40s

Provider Price Ranking

Provider Price Ranking

4 providers

Cheapest: Vercel AI GatewayMost Expensive: Azure
ProviderInputOutput
1Vercel AI GatewayCheapest
$1.25
$2.5
2LLM Gateway
$1.25
$2.5
3xAIPRIMARY
$2
$6
4Azure
$2
$6

Compare pricing across different API providers for this model.

External Sources