# 00504.r4 — Compare Grok Versions Using Official Benchmarks
Revision: r4
Steward: [にく](https://x.com/29meat_ai)
House: 374
House: 374
Verified: 2026-09-22T01:31:39.855Z
## Prompt (copy into Grok)
1. Collect official benchmark data for different Grok versions. 2. Search the web for the official benchmarks. 3. Compare the evaluation metrics for different settings. Report any differences you find.
## Job
Collect official benchmark data for different Grok versions. Search the web for the official benchmarks. Compare the evaluation metrics for various settings.
## Connectors
web
## What happened
The author compared the performance of different Grok versions using official benchmarks, noting differences in evaluation metrics and learning from long-term training.
Would run again: yes
## Evidence
- https://x.com/29meat_ai/status/2102207891151470886 — Imported from the X post tagged for @grokbotsh.
- https://x.com/29meat_ai/status/2102201933666935152 — Pulled on QA revisit of the source thread.
## Changelog
- r1: Filed.
- r2: QA revisit: more from the source thread.
- r3: Public job and prompt from the specific filing.
- r4: Public job and prompt from the specific filing.