跳轉到主要內容

Grok 4.1 Fast (Reasoning)

SpaceXAIGrokProprietary

描述

Grok 4.1 Thinking (code name: quasarflux) is the reasoning variant of Grok 4.1, bringing significant improvements to the real-world usability of Grok. The model is exceptionally capable in creative, emotional, and collaborative interactions. It is more perceptive to nuanced intent, compelling to speak with, and coherent in personality, while fully retaining the razor-sharp intelligence and reliability of its predecessors. Grok 4.1 Thinking uses thinking tokens for deeper reasoning, achieving state-of-the-art performance on blind human preference evaluations. The model features reduced hallucinations compared to previous versions, with significant improvements in factual accuracy for information-seeking prompts. To achieve this, xAI used large scale reinforcement learning infrastructure to optimize style, personality, helpfulness, and alignment, developing new methods that use frontier agentic reasoning models as reward models to autonomously evaluate and iterate on responses at scale. Grok 4.1 Thinking includes comprehensive safety mitigations including refusal training, input filters for restricted knowledge, and adversarial robustness measures. The model demonstrates strong refusal rates on harmful queries (7% answer rate on chat refusals, 14% on agentic refusals) and improved honesty training to reduce deception.

發布日期
2025-11-19
參數規模
—
上下文長度
—
支援模態
image, text

能力雷達圖

50
general
40
coding
21
reasoning
26
science估算
40
agents
80
multimodal

缺少專門科學評測時,Science 由 LLM Stats 科學得分或推理能力估算。

排行榜排名

暫無排名資料

基準測試分數 (LLM Stats)

(LLM Stats (zeroeval))

Biology

BioLP-Bench37.0%自報

Chat

EQ-Bench1586.00 / 2000自報
LMArena Text Leaderboard1483.00 / 2000自報

Creativity

Creative Writing v386.1%自報

Safety

WMDP84.0%自報
ProtocolQA79.0%自報
Virology Capabilities Test61.0%自報
MASK51.0%自報
CloningScenarios46.0%自報
CyBench39.0%自報
FigQA34.0%自報

AA 評測指數

(Artificial Analysis)

暫無 AA 評測資料

LLM Stats 分類評分

(LLM Stats (zeroeval))
Creativity
90
Writing
90
General
50
Safety
40
Agents
40
Code
40
Healthcare
30
Vision
30

定價

輸入價格$0.2 / 1M tokens
輸出價格$0.5 / 1M tokens
混合價格(3:1)$0.275 / 1M tokens

速度

暫無速度資料

供應商價格排行

供應商價格排行

8 個供應商

最便宜: Jiekou.AI最貴: Azure
供應商輸入輸出
1Jiekou.AI最便宜
$0.19
$0.475
2SpaceXAI主要
$0.2
$0.5
3302.AI
$0.2
$0.5
4Helicone
$0.2
$0.5
5FrogBot
$0.2
$0.5
6Azure Cognitive Services
$0.2
$0.5
7DevPass (LLM Gateway)
$0.2
$0.5
8Azure
$0.2
$0.5

比較該模型在不同 API 供應商之間的定價。

外部連結