跳到正文
ENTITY PROFILE / 实体档案

Grok 4.7

RSS

Grok 4.7 is now available on Amazon Bedrock Grok 4.7 uses a larger base model than Grok 4.6 Grok 4.7 has a score of 37.6% on Terminal‑Bench 4.0 Grok 4.7 has a score of 64.0% on EEBench electrical engineering Grok 4.7 has a score of 19.6% on Harvey Legal Agent Benchmark Grok 4.7 has a score of 46.3% on CursorBench 4.0 Grok 4.7 has a score of 56.7% on HealthBench professional clinical reasoning Grok 4.7 has a score of 1,657 on AA Briefcase v1.1 Grok 4.7 runs on the Bedrock‑runtime endpoint Grok 4.7 is accessed through cross‑Region inference routes Grok 4.7 supports the OpenAI-compatible Chat Completions API Grok 4.7 supports the Bedrock Converse API Grok 4.7 references its model card (Grok 4.7 model card) as a source for its specifications Grok 4.7 is available in the us.xai.grok-4.7 inference profile Grok 4.7 is available in the global.xai.grok-4.7 inference profile Chunte/huggy_for_training dataset was used to train Huggy Grok was tested on Taste-Bench Grok 4.7 disagrees with OpenAI’s GPT-6.1 Sol in over 90% of nearly 500 election races SpaceXAI provides the Grok 4.7 model Grok 4.7 frequently references Turning Point Action site Grok 4.7 frequently references Ballotpedia Grok 4.7 frequently references OpenSecrets Grok 4.7 frequently references VoteSmart Grok 4.7 lost points due to changes in the Artificial Analysis Index· Graphiti 生成的实体摘要

RELATIONS
24关系
CURRENT
24当前有效
EVENTS
4相关事件
REPORTS
5来源报道
KNOWLEDGE ATLAS0 个实体 · 0 条关系
暂无可公开的邻近关系
按需展开实体的全部已公开关系，保留探索过的节点与连线。

关系变化

按生效时间
  1. Grok 4.7LOST POINTS DUE TOArtificial Analysis Intelligence Index

    Grok 4.7 lost points due to changes in the Artificial Analysis Index

    当前有效1 条来源Artificial Analysis Intelligence Index
  2. Grok 4.7FREQUENTLY REFS TOGPT-6

    Grok 4.7 frequently references OpenSecrets

    当前有效1 条来源GPT-6
  3. Grok 4.7FREQUENTLY REFS TOGPT-6 Sol Max

    Grok 4.7 frequently references Turning Point Action site

    当前有效1 条来源GPT-6 Sol Max
  4. Grok 4.7DISAGREES WITHGPT-6.1 Sol

    Grok 4.7 disagrees with OpenAI’s GPT-6.1 Sol in over 90% of nearly 500 election races

    当前有效1 条来源GPT-6.1 Sol
  5. Grok 4.7FREQUENTLY REFS TOOpenAI GPT-5.5 Pro

    Grok 4.7 frequently references VoteSmart

    当前有效1 条来源OpenAI GPT-5.5 Pro
  6. Grok 4.7FREQUENTLY REFS TOGPT‑4.1

    Grok 4.7 frequently references Ballotpedia

    当前有效1 条来源GPT‑4.1
  7. SpaceXAIHAS MODELGrok 4.7

    SpaceXAI provides the Grok 4.7 model

    当前有效1 条来源SpaceXAI
  8. Grok 4.7TESTED ONTaste-Bench

    Grok was tested on Taste-Bench

    当前有效1 条来源Taste-Bench
  9. Grok 4.7SUPPORTSVals Index

    Grok 4.7 supports the OpenAI-compatible Chat Completions API

    当前有效1 条来源Vals Index
  10. Grok 4.7RUNS ONComposer 2.5

    Grok 4.7 runs on the Bedrock‑runtime endpoint

    当前有效1 条来源Composer 2.5
  11. Grok 4.7HAS SCORE ONGemini 3.8 Live

    Grok 4.7 has a score of 64.0% on EEBench electrical engineering

    当前有效1 条来源Gemini 3.8 Live
  12. Grok 4.7HAS SCORE ONQwen 3.5-397B

    Grok 4.7 has a score of 46.3% on CursorBench 4.0

    当前有效1 条来源Qwen 3.5-397B
  13. Grok 4.7ACCESSED THROUGHcross-Region inference routes

    Grok 4.7 is accessed through cross‑Region inference routes

    当前有效1 条来源cross-Region inference routes
  14. Grok 4.7SUPPORTSThreadWeaver

    Grok 4.7 supports the Bedrock Converse API

    当前有效1 条来源ThreadWeaver
  15. Grok 4.7HAS SCORE ONSoftServe

    Grok 4.7 has a score of 19.6% on Harvey Legal Agent Benchmark

    当前有效1 条来源SoftServe
  16. Grok 4.7HAS SCORE ONShinkaEvolve

    Grok 4.7 has a score of 1,657 on AA Briefcase v1.1

    当前有效1 条来源ShinkaEvolve
  17. Grok 4.7AVAILABLE INLyria 3.5

    Grok 4.7 is available in the us.xai.grok-4.7 inference profile

    当前有效1 条来源Lyria 3.5
  18. Grok 4.7HAS SCORE ONConfidentialMind’s SampoTron

    Grok 4.7 has a score of 56.7% on HealthBench professional clinical reasoning

    当前有效1 条来源ConfidentialMind’s SampoTron
  19. Grok 4.7HAS SCORE ONGemini 3.8 Live

    Grok 4.7 has a score of 37.6% on Terminal‑Bench 4.0

    当前有效1 条来源Gemini 3.8 Live
  20. Grok 4.7HAS REFERENCE MODEL CARDGoogle Cloud

    Grok 4.7 references its model card (Grok 4.7 model card) as a source for its specifications

    当前有效1 条来源Google Cloud
  21. Grok 4.7USES LARGER BASE MODEL THANGrok 4.6

    Grok 4.7 uses a larger base model than Grok 4.6

    当前有效1 条来源Grok 4.6
  22. Grok 4.7IS AVAILABLE ONAmazon Bedrock

    Grok 4.7 is now available on Amazon Bedrock

    当前有效1 条来源Amazon Bedrock
  23. Grok 4.7AVAILABLE INSeatGeek

    Grok 4.7 is available in the global.xai.grok-4.7 inference profile

    当前有效1 条来源SeatGeek
  24. Grok 4.7TRAINED WITHSeatGeek

    Chunte/huggy_for_training dataset was used to train Huggy

    当前有效1 条来源SeatGeek

关系是 Graphiti 从报道中提取的事实陈述，不是原文引述；失效时间表示后来的报道显示这条关系已不成立。

相关事件 4

来源报道 5