Model
Model explorer

Grok-2

CLOSED
SpaceXAI · Grok 2 family · released Aug 13, 2024

Multimodal frontier beta with FLUX.1-powered image generation on X; an early build ("sus-column-r") ranked #2-3 on the LMSYS/Chatbot Arena leaderboard pre-release, matching GPT-4o (did not reach #1).

ReasoningCodingVisionFunction callingTool useAgentic
1228.2
Elo · rank #177
Parameters
Undisclosed
Active params
Undisclosed
Context
131K tokens
Architecture
Mixture-of-Experts (size undisclosed by xAI; third-party estimates unverified)
License
Proprietary
Languages
—
API price (in/out)
$2 / $10
Modalities
text · vision
Benchmark results
Bar shows position within the tracked field; marker = field best
DocVQAVision93.6%#13
best: Qwen2-VL-72B · 96.5%
GPQA DiamondReasoning56.0%#174
best: GPT-5.6 · 94.6%
HumanEvalCoding88.4%#27
best: Claude Opus 4.5 · 99.4%
MATH-500Math76.1%#77
best: GPT-5 · 99.4%
MathVistaVision69.0%#29
best: Seed 2.1 Pro · 90.7%
MMLU-ProKnowledge75.5%#74
best: Claude Fable 5 · 91.5%
MMLUKnowledge87.5%#30
best: OpenAI o3 · 92.9%
MMMUVision66.1%#65
best: Claude Fable 5 · 89.3%
Run it locally
Closed weights — available via API only. No local deployment.
Input / M tok
$2
Output / M tok
$10
Grok 2 family
Elo progression across releases
Grok-2Aug 20241228.2Grok-2 miniAug 20241129.5
Resources
DOCSdocs.x.aiAPIAvailable via SpaceXAI API
API price $2/$10 · each benchmark row carries its own source badge (see methodology)