Grok 4 Leads Select Benchmarks as xAI, Anthropic Race Scales Up
Embed this grade
Why this grade: Graded B+: core historical timeline, Colossus specs, Grok 4 benchmarks, and funding figures align closely with public records; minor spelling errors, optimistic unreleased Grok 5 projections, and one unverified Colossus tenancy claim reduce score
Why this lean: Neutral presentation balancing both companies' philosophies, strengths, and enterprise positions without favoring one side
Disagree with this grade or political lean?
Tell us why. Your note is reprocessed through the same grading logic; if the output is still off the report is removed, and if it holds up it stays.
Topics in this report
Summary
The broadcast traces the origins of Anthropic and xAI from OpenAI alumni, details the Colossus supercluster buildout in Memphis, compares Grok 4 and Claude 4 family capabilities on benchmarks, and discusses Grok 5 expectations. It covers training approaches, enterprise use cases, real-time data advantages, and the broader capital race. Sourcing relies on company announcements, benchmark leaderboards, and reported funding rounds; no named external guests appear.
Editorial Assessment
The video provides a detailed, largely accurate overview of the competitive landscape as of mid-2026, correctly citing Grok 4's strong performance on Humanity's Last Exam, GPQA, and AIME while noting Claude's edge in coding and legal tasks. Colossus scale and timelines match official and press reports. Minor factual slips include misspelled names and an unconfirmed claim that Anthropic rents Colossus GPUs. Grok 5 details remain speculative projections. Viewers miss primary source links for some benchmark numbers and updated environmental compliance status at the Memphis site.
Key Moments
Dario and Daniela Amodei left OpenAI in late 2020 to found Anthropic in 2021 over safety focus
Confirmed by OpenAI's December 2020 announcement and subsequent reporting on the split
Colossus built in 122 days with 100k GPUs operational July 2024, expanded to 200k then 550k GPUs by late 2025
Matches xAI statements and contemporaneous coverage of Memphis facility timeline
Grok 4 scores ~44.4% standard / 50.7% heavy on Humanity's Last Exam, 87-88% GPQA, 95% AIME
Directly corroborated by xAI's July 2025 Grok 4 announcement and benchmark summaries
xAI valued at ~$230B after $20B round in early 2026; Anthropic reached $380B then neared $1T
Aligns with official funding announcements from both companies in 2026
Grok 5 unreleased as of mid-2026, anticipated on expanded Colossus with potential 6T parameters
Release delayed past Q2 2026 per reports; parameter count remains unconfirmed speculation
Notable Concerns
- Minor name spelling inconsistencies (Daario instead of Dario)
- Unverified claim of Anthropic tenancy at Colossus facility
Sources Consulted
- Anthropic Co-Founder Dario Amodei Explains Why He Left OpenAI
- Anthropic’s CEO says why he quit his job at OpenAI to start a competitor
- Colossus: The World's Largest AI Supercomputer - xAI
- Elon Musk updates xAI's Colossus 2 in Memphis
- Shocking Scores of Musk's Grok 4 Leaked
- Grok 4: #1 on Humanity’s Last Exam
- Humanity's Last Exam - Center for AI Safety
- ALTSS Hot 30 AI Companies — Q1 2026
- Trading Places Podcast - Practical VC
- Analysis: The Fintech Blueprint AI & Robotics Report Q1 2026