Private AI
Grok 4.20 Multi-Agent is tuned for collaborative agentic workflows while keeping the same 2M-token context window and multimodal support. Content policy rejections can still be charged: xAI may pass through a $0.05 moderation-failure fee, or a $0.055 usage-guidelines violation fee, depending on which rejection upstream returns.
Added Mar 31, 2026
Context Window
2.0M
Max Output
131.1K
Avg output tokens (7d)
8.1K tokens
Input Price (Auto)
$2.00/1M
Output Price (Auto)
$6.00/1M
Cache Read (Auto)
$1.00/1M
Capabilities
Performance metrics and benchmarks
Sourced from LMArena.
Arena Score
1470.0
Overall Rank
#28 / 376
Votes
56,758
Confidence Interval
1466.0 - 1473.9
Coding
#37 / 371
15,784 votes
1509.6
Math
#47 / 364
3,041 votes
1451.4
Longer Query
#51 / 354
23,497 votes
1458.4
Creative Writing
#25 / 374
9,452 votes
1448.2
Instruction Following
#51 / 376
18,672 votes
1442.7
Hard Prompts
#39 / 376
36,598 votes
1483.4
Auto routing is available for this model. Explicit provider selection is not available.
Loading provider options…
French
#18 / 260
1,974 votes
1492.7
Polish
#18 / 203
1,198 votes
1480.9
German
#20 / 279
921 votes
1472.0
Russian
#20 / 338
5,917 votes
1480.6
Korean
#25 / 250
908 votes
1430.4
Non English
#25 / 376
30,071 votes
1460.2
Spanish
#25 / 261
1,736 votes
1464.1
Industry Software And It Services
#28 / 376
22,347 votes
1502.7
Exclude Ties
#29 / 376
42,935 votes
1476.7
Industry Entertainment And Sports And Media
#30 / 374
11,979 votes
1441.8
English
#31 / 376
26,685 votes
1472.9
Industry Legal And Government
#34 / 348
4,456 votes
1474.5
Industry Medicine And Healthcare
#34 / 345
4,170 votes
1482.5
Multi Turn
#34 / 374
9,286 votes
1474.4
Industry Life And Physical And Social Science
#36 / 374
9,223 votes
1480.6
Chinese
#40 / 346
2,918 votes
1499.6
Industry Writing And Literature And Language
#41 / 375
13,668 votes
1445.9
Industry Business And Management And Financial Operations
#43 / 369
11,246 votes
1457.3
Hard Prompts English
#48 / 375
18,070 votes
1480.5
Industry Mathematical
#48 / 357
3,031 votes
1456.1
Japanese
#49 / 242
508 votes
1412.1
Expert
#50 / 324
5,371 votes
1479.1
Published 2026-07-16 · Matched as grok-4.20-multi-agent-beta-0309
LMArena Dataset