SpaceXAI: Grok 4.20 Multi-Agent

Grok 4.20 Multi-Agent is a variant of SpaceXAI’s Grok 4.20 designed for collaborative, agent-based workflows. Multiple agents operate in parallel to conduct deep research, coordinate tool use, and sy…

x-ai/grok-4.20-multi-agent-20260309

Grok 4.20 Multi-Agent is a variant of SpaceXAI’s Grok 4.20 designed for collaborative, agent-based workflows. Multiple agents operate in parallel to conduct deep research, coordinate tool use, and synthesize information across complex tasks.

Reasoning effort behavior: - low / medium: 4 agents - high / xhigh: 16 agents

Specifications & capabilities

Input
text, image, file
Output
text
Context window · tokens
2000000
Knowledge cutoff
2025-09-01
Published model ID
grok-4.20-multi-agent-0309
Capabilities
fileInput, multimodal, promptCaching, reasoning, structuredOutputs, toolCalling, visionInput, webSearch
Description source ↗

Evaluations

1–20 / 107

LMArena

1,265.571

source id
lmarena
leaderboard name
LMArena Vision Entity Recognition
dimension
multimodal
metric
Arena rating
rank
15
sample size
167
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1222.707332308566
confidence upper
1308.434670228274
Original result ↗

LMArena

1,206.975

source id
lmarena
leaderboard name
LMArena Search Style Control Overall
dimension
search
metric
Arena rating
rank
15
sample size
121069
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1201.3078042125371
confidence upper
1212.6425354361554
Original result ↗

LMArena

1,438.721

source id
lmarena
leaderboard name
LMArena Text Factuality German
dimension
factuality
metric
Arena rating
rank
16
sample size
620
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1416.782947035133
confidence upper
1460.6596468038083
Original result ↗

LMArena

1,227.194

source id
lmarena
leaderboard name
LMArena Vision Style Control Entity Recognition
dimension
multimodal
metric
Arena rating
rank
18
sample size
167
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1184.0244578057443
confidence upper
1270.3636435224767
Original result ↗

LMArena

1,205.156

source id
lmarena
leaderboard name
LMArena Search Overall
dimension
search
metric
Arena rating
rank
18
sample size
121069
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1199.6103728611702
confidence upper
1210.701589580251
Original result ↗

LMArena

1,198.056

source id
lmarena
leaderboard name
LMArena Search Factuality Overall
dimension
search
metric
Arena rating
rank
20
sample size
121017
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1193.7860767053392
confidence upper
1202.3255805027552
Original result ↗

LMArena

1,280.034

source id
lmarena
leaderboard name
LMArena Vision Humor
dimension
multimodal
metric
Arena rating
rank
27
sample size
925
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1259.8844498689898
confidence upper
1300.1829200456089
Original result ↗

LMArena

1,438.537

source id
lmarena
leaderboard name
LMArena Text Factuality Polish
dimension
factuality
metric
Arena rating
rank
29
sample size
890
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1419.5266940741615
confidence upper
1457.5471079462252
Original result ↗

LMArena

1,428.341

source id
lmarena
leaderboard name
LMArena Text Factuality Japanese
dimension
factuality
metric
Arena rating
rank
29
sample size
428
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1405.4633282292416
confidence upper
1451.2188836987984
Original result ↗

LMArena

1,281.515

source id
lmarena
leaderboard name
LMArena Vision Creative Writing Vision
dimension
multimodal
metric
Arena rating
rank
31
sample size
1666
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1266.2200613065136
confidence upper
1296.8099087666203
Original result ↗

LMArena

1,266.435

source id
lmarena
leaderboard name
LMArena Vision Style Control Creative Writing Vision
dimension
multimodal
metric
Arena rating
rank
31
sample size
1666
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1251.066172493223
confidence upper
1281.8047346383562
Original result ↗

LMArena

1,263.205

source id
lmarena
leaderboard name
LMArena Vision Style Control Humor
dimension
multimodal
metric
Arena rating
rank
33
sample size
925
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1242.949553951779
confidence upper
1283.4602783625041
Original result ↗

LMArena

1,479.417

source id
lmarena
leaderboard name
LMArena Text Style Control Polish
dimension
style
metric
Arena rating
rank
36
sample size
1314
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1463.182818667392
confidence upper
1495.6521093892857
Original result ↗

LMArena

1,491.912

source id
lmarena
leaderboard name
LMArena Text Style Control French
dimension
style
metric
Arena rating
rank
37
sample size
2234
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1477.7820542241134
confidence upper
1506.0429451352074
Original result ↗

LMArena

1,435.718

source id
lmarena
leaderboard name
LMArena Text Style Control Korean
dimension
style
metric
Arena rating
rank
38
sample size
998
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1415.9576606598935
confidence upper
1455.47831073192
Original result ↗

LMArena

1,482.034

source id
lmarena
leaderboard name
LMArena Text Factuality French
dimension
factuality
metric
Arena rating
rank
39
sample size
1769
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1467.8477465586448
confidence upper
1496.2211057547747
Original result ↗

LMArena

1,478.534

source id
lmarena
leaderboard name
LMArena Text Style Control Russian
dimension
style
metric
Arena rating
rank
40
sample size
7076
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1470.778227766999
confidence upper
1486.2898526430147
Original result ↗

LMArena

1,474.795

source id
lmarena
leaderboard name
LMArena Text Style Control German
dimension
style
metric
Arena rating
rank
41
sample size
1056
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1455.4868351288274
confidence upper
1494.1039330085641
Original result ↗

LMArena

1,442.85

source id
lmarena
leaderboard name
LMArena Text Factuality Creative Writing
dimension
factuality
metric
Arena rating
rank
44
sample size
9246
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1436.827816606095
confidence upper
1448.8723214526012
Original result ↗

LMArena

1,460.617

source id
lmarena
leaderboard name
LMArena Text Style Control Non English
dimension
style
metric
Arena rating
rank
45
sample size
34959
evaluated at
2026-11-09T00:00:00.000Z
source model id
Grok 4.20 Multi-Agent Beta (0309)
confidence lower
1456.0734178046355
confidence upper
1465.1605170554055
Original result ↗

Published prices

21–22 / 22

grok-4.20-multi-agent-0309 ↗

$10.00

Inference provider
xAI
Platform
xAI
Unit
output_tokens_per_million
Currency
USD
Billing conditions
billing_rule
Only when response confirms service_tier=priority
service_tier
Priority
input_tokens_gte
200000
Billing label
Output / 1M tokens

grok-4.20-multi-agent-0309 ↗

$5.00

Inference provider
xAI
Platform
xAI
Unit
output_tokens_per_million
Currency
USD
Billing conditions
billing_rule
Only when response confirms service_tier=priority
service_tier
Priority
input_tokens_lt
200000
Billing label
Output / 1M tokens

Provider specifications & access

1–2 / 2

grok-4.20-multi-agent ↗

source model id
grok-4.20-multi-agent
hosting
gateway
platforms
  • Cheaper Inference
context length
2000000
max output tokens
2000000
architecture
input_modalities
  • text
  • image
output_modalities
  • text
capabilities
  • reasoning
  • structuredOutputs
  • toolCalling
  • visionInput
  • webSearch

Grok 4.20 Multi-Agent Beta ↗

source model id
grok-4.20-multi-agent-0309
hosting
xai-hosted
platforms
  • xAI
context length
1000000
architecture
input_modalities
  • text
  • image
output_modalities
  • text
capabilities
  • toolCalling
  • structuredOutputs
  • reasoning

Safety assessments

1–1 / 1

Foundation Model Transparency Index 2025 ↗

14 / 100

source id
stanfordFMTI
canonical model id
x-ai/grok-4.20-multi-agent-20260309
taxonomy
transparency
hazard category
transparency
hazard name
Organization transparency
policy area
xAI
assessment type
organization-level disclosure
metric
disclosure_score
result label
14 / 100
sample size
100
updated at
2026-11-09T18:08:05.227Z

Source specification observations

1–4 / 4

openRouter ↗

field
context_length
value
2000000
normalized value
2000000
scope
source-model
source id
openRouter
confidence
1

xAI ↗

field
context_length
value
1000000
normalized value
1000000
scope
source-model
source id
xAI
confidence
1

openRouter ↗

field
knowledge_cutoff
reported field
knowledge_cutoff
value
2025-09-01
normalized value
2025-09-01
precision
day
scope
source-model
source id
openRouter
confidence
1

openRouter ↗

field
max_output_tokens
value
1800000
normalized value
1800000
scope
source-model
source id
openRouter
confidence
1

Original sources

1–7 / 7

OpenRouter · Model page ↗

xAI · Model documentation ↗

api.cheaperinference.com · Source snapshot ↗

GitHub · Catalog ↗

LMArena · Catalog ↗

OpenRouter · Catalog ↗

xAI · Catalog ↗

Source links

1–5 / 5

Explore more models

Models

Loading
EVIDENCE / COMPARISON

Benchmarks

Explore every benchmark

Loading

These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.

Install TensorTrace