Grok-2
About
xAI Grok-2 large language model with extended context for reasoning.
Capabilities
VisionMultimodalReasoningFunction CallingTool UseJSON ModeCode Execution
Benchmark Scores(2)
| Benchmark | Score | Version | Source |
|---|---|---|---|
| HumanEval | 88.4 | pass@1 | Open LLM Leaderboard, xAI official |
| Massive Multitask Language Understanding | 87.5 | 5-shot | Open LLM Leaderboard, xAI official |
Specifications
FamilyGrok
Released2024-08-01
Context128K
ArchitectureDecoder Only
Specializationgeneral
Trainingfinetuning