Skip to content

Grok 4.20

xAIProprietaryReasoningProvisionalReleased Mar 10, 2026Input: text, image, file
Overall score
57.4
Rank 61 of 222 · 47.7–67.2
Context window
2M
2,000,000 tokens
Price per 1M tokens
$1.25 in · $2.50 out
Source: BenchLM
Blended price
$1.56
3:1 input:output tokens
Output speed
—
Not measured yet

Category scores

Scores per capability area with the model's rank among ranked models in that category.

  1. 1Knowledge59.2
  2. 2Multimodal#4853.4
  3. 3Coding49.3
  4. 4Reasoning44.9
  5. 5Agentic & tool use32.1
Higher is better. #N is the model's rank in that category, where it is eligible to be ranked.

Benchmark results (18)

As reported in system cards, papers and public evaluations. Position compares this model with every other model reporting the same benchmark.

Multimodal

Design Arena Website1231#50 of 91
MMMU-Pro75.2#32 of 48
CharXiv60.9#41 of 42
SimpleVQA57.4#14 of 17
ERQA54.1#14 of 16
MedXpertQA (MM)65.8#6 of 8

Agentic & tool use

Gert Labs38.36#42 of 52
MazeBench0#20 of 42
Terminal-Bench 2.047.1#32 of 39
DeepSearchQA62.8#19 of 20

Coding

SWE-bench Verified76.7#37 of 98
SWE-bench Pro51.8#57 of 69
LiveCodeBench Pro74.2#4 of 6

Knowledge

HLE w/o tools31.6#29 of 43
HealthBench Hard20.3#11 of 12
MedXpertQA (Text)50.2#5 of 5

Reasoning

ARC-AGI-253.3#32 of 41
ARC-AGI-30.1#22 of 22

Pricing by source

USD per 1M tokens.

SourceInputOutputCached in
List price (BenchLM)$1.25$2.50$0.20
OpenRouter x-ai/grok-4.20$1.25$2.50$0.20

xAI's pricing page, read September 25, 2026, lists grok-4.20-0309-reasoning at $1.25 input / $0.20 cached input / $2.50 output per million tokens below 200K prompt tokens, with a 1M context window. Prompts of 200K tokens or more bill every token at $2.50 / $0.40 / $5.00. The grok-4.20-0309-non-reasoning variant lists at the same rates. BenchLM previously recorded $2.00 / $6.00 for this row.

Arena ratings

Human-preference ratings from LMArena. Best-rated variant shown.

WebDev (coding)(Reasoning, 0309)1376#107 · 15.4K votes
Vision(Reasoning, 0309)1262#55 · 26.8K votes

Speed

Measured by us through OpenRouter (median of trials).

Not measured yet.

Related models

Similar score