Grok Beta: Complete Benchmark Performance Guide

xai beta 2023-11-04 Proprietary

In-depth analysis of Grok Beta performance on MMLU, HumanEval, GSM8K, MATH, BBH benchmarks, including pricing and use cases.

Key Specifications

Vendorxai
Versionbeta
Release Date2023-11-04
Context Window8192 tokens
Input Modalitiestext
Output Modalitiestext
LicenseProprietary
Documentationhttps://docs.x.ai/

Benchmark Performance

Benchmark Score Unit Evaluated At Notes Source
MMLU 51 % 2023-11-04 5-shot view
HUMANEVAL 31 pass@1 2023-11-04 — view
GSM8K 48.7 % 2023-11-04 0-shot CoT view
MATH 19.1 % 2023-11-04 0-shot CoT view
BBH 50.2 % 2023-11-04 3-shot CoT view
GPQA 20.6 % 2023-11-04 0-shot view
IFEVAL 56.2 % 2023-11-04 prompt_strict view
ARC 81.7 % 2023-11-04 challenge view
MUSR 43.7 % 2023-11-04 0-shot view
WINOGRANDE 69.2 % 2023-11-04 0-shot view

Pricing

Tier Price Currency
Input$1 / MtokUSD
Output$5 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://x.ai/api · as of 2023-11-04

Compliance

  • Data Residency: US
  • SOC2: ✗
  • HIPAA: ✗
  • GDPR: ✗
  • ISO 27001: ✗

Grok Beta

Model Overview

xAI Grok Beta 早期预览版, 8K 上下文, 集成 X 平台实时数据, 幽默风格对话能力突出。

Core Specifications

Vendor Version Release Date Context Window Input Modalities Output Modalities License
Xai beta 2023-11-04 8K text text Proprietary

Benchmark Performance

Benchmark Score Unit Notes
MMLU (Massive Multitask Language Understanding) 51.0 % 5-shot
HumanEval 31.0 pass@1 —
GSM8K (Grade School Math 8K) 48.7 % 0-shot CoT
MATH 19.1 % 0-shot CoT
BBH (BIG-Bench Hard) 50.2 % 3-shot CoT
GPQA 20.6 % 0-shot
IFEval 56.2 % prompt_strict
ARC 81.7 % challenge
MUSR 43.7 % 0-shot
WinoGrande 69.2 % 0-shot

Pricing

Input Output Cache Read Cache Write
— — — —

per million tokens

Strengths

  • Reliable general-purpose model.

Weaknesses

  • MMLU 51.0, weak knowledge reasoning.
  • HumanEval 31.0, coding weak.
  • Proprietary, not self-hostable.
  • Context window 8K is limited.

Use Cases

  • General chat and Q&A

References