GPT-5.6 about to DROP
Credibility score: 44/100 — Mixed Credibility. Several questionable claims detected. Watch with healthy skepticism.
BSmeter analyzed "GPT-5.6 about to DROP" and rated it 44/100 for credibility (a BS score of 56/100 — mixed credibility), on 2026-06-02. Its weakest claim — "Anthropic filed confidentially for IPO at nearly $8 trillion valuation" — scored 20/100 and was flagged as sketchy. 11 claims were checked against the video transcript. Scores are produced by BSmeter's AI analysis of the transcript, not independent human verification.
Claims analyzed
Video opens with highlight reel preview — Just Vibes (50/100)
Bold preview — let's see if the video delivers on these promises.
Anthropic filed confidentially for IPO at nearly $8 trillion valuation — Sketchy (20/100)
Re-scored to 20/100 after evidence check.
Sources: San Francisco AI giant Anthropic files for IPO after $965 billion valuation, Anthropic files to go public in a potentially trillion-dollar debut | CNN Business, The AI IPO race is on: Anthropic confidentially files after surpassing OpenAI in valuation - San Francisco Business Times
Opus 4.8 built entire simulation in ultra mode — Unverifiable (50/100)
No public evidence Opus 4.8 exists — naming feels off.
Claude Opus 4.8 loses to GPT-5.5 on software engineering benchmark — Unverifiable (50/100)
No benchmark name, numbers, or source given — pure assertion
Benchmark skipped Claude's 'ultra code' mode entirely — Opinion (50/100)
Fair critique of missing test condition — still speculation
If Claude ultra code loses, Anthropic is falling behind OpenAI — Opinion (50/100)
One benchmark ≠ company trajectory — massive leap
RKGI benchmark tests fluid intelligence in AI models — OK (55/100)
Sounds plausible but no public info on what RKGI actually is.
Opus 4.8 set SOTA on ARC-AGI 3 at 1.5% — Dubious (40/100)
ARC-AGI 3 leaderboard doesn't list any Opus 4.8 model at 1.5%.
Opus 4.8 scored 1.5% on some benchmark while others sit at 0.5% or below — Unverifiable (50/100)
No benchmark named — can't check the actual numbers
GPT-5.6 and GPT-5.6 Pro spotted in OpenAI Codex backlogs, possible release tomorrow or Tuesday — Dubious (45/100)
Rumors are real but evidence is thin — no official confirmation anywhere.
OpenAI released GPT-5.5 in April, so GPT-5.6 coming in June means frequent rolling updates — Sketchy (25/100)
GPT-5.5 in April? No evidence of that release either — timeline is invented 🗓️💀
See the full analysis with sources and timestamps →