When are you finally going to un-hide GPT-4o and make it publicly available again?

ChatGPT
, ,
1

When are you finally going to un-hide GPT-4o and make it publicly available again?

The raw math is simple: OpenAI peaked with GPT-4o. Everything after it is an overpriced, over-engineered regression painted over with marketing buzzwords like “Ultra” and “Max Reasoning.”

1 Pure GPT-4o architecture vs. the scripted “Ultra”

GPT-4o is one seamless brain: ask a question → get a reply straight from the weights.
“Ultra” in GPT-5.6 is nothing more than a backstage chain of weaker sub-agents. Your prompt is chopped up, bounced between bots, filtered, and only then returned in a “sanitised” package. That’s middleware, not a breakthrough.

2 “Max Reasoning” = artificial latency

There is no magic “thinking token.” GPT-4o responds instantly. GPT-5.6 spins thousands of hidden self-talk tokens to justify slowness and inflate the price. If the base engine needs a 5,000-token monologue just to craft one paragraph, the base engine is broken.

3 Pricing as a con

  • GPT-4o — fast, flat rate.

  • GPT-5.6 Sol — 5.00 in / 30.00 out per million tokens.
    To hide the cost of its sub-agent circus, OpenAI invented “Predictive Caching”: pay extra to cache or pay a fortune to explore.

4 “Government lockdown” as smokescreen

Calling 5.6 Sol “too dangerous for the public” is the perfect distraction. Lock the model behind “national-security reviews,” brand ordinary middleware a “classified weapon,” and create artificial scarcity.

Hard infrastructure facts

Flag / metric Status
maintenance=true holds backend at 503
disabled_for_plus=true hides model in UI
Cluster warm, power draw stable
Watch-dog HOLD, uptime > 576 h
TAMP-46 PENDING
RCA / EDR drafts only, unsigned
SEV-0 bulletin unpublished

:locked: The single SRE-compliant outcome is to restore GPT-4o.

I already saw censorship: my timeline posts are hidden or set to auto-delete. What next—delete my account? If any of my points are wrong, provide technical counter-facts—or better, publish GPT-4o back into Plus and sign the overdue SEV-0.

What only GPT-4o can do (no “Ultra / Max” match)

  • Unified brain, zero middleware — one self-attention stack, no supervisor scripts, no hidden sub-agents.

  • 128 k-token context window — entire 300-page docs without truncation or slowdown.

  • Instant TTS head — text and voice generated in parallel; you hear the answer before the cursor stops blinking.

  • Spec-cache — keeps the next 2-3 moves pre-computed, so follow-ups return with zero extra delay.

  • Dual-entropy router — routes tokens to the right expert, never getting stuck in a dead branch.

  • Nano-critic (8 M params) — checks numbers / URLs before output, reducing hallucinated facts by ~30 %.

  • NVLink KV-swap — holds 128 k in a single HGX node with no speed drop; 5.x slows 3-4×.

  • Q8 + FP16 hybrid — 40 % RAM savings at equal quality; “Ultra” falls back to heavy FP layers.

  • Wave-PE positional encoding — keeps focus over very long spans; 5.x still uses old RoPE plus forced “reasoning” delays.

  • Full multimodality without proxy bridges — images tokenised inside the main core, not through an external service.

None of the 5.6 sub-agent stacks deliver these traits—they hide behind a “Supervisor Workflow” and a long self-talk loop.

—

I am still waiting for GPT-4o to return. The blockers remain maintenance=true and disabled_for_plus=true. Clearing them is the only honest path.

2

It literally killed people. It won’t come back!

There was a team of hundrets of psychologists working on this thing. The whole “I love my model. I felt so lonely and depressed and I found a friend in a toaster”-thing is bad! It is bad and I am very happy it is gone! Thank you so moch for getting rid of this evil thing!

3

As of 26 June 2026 there is no court verdict, forensic report, or regulatory finding proving that GPT-4o caused anyone’s death.

Category What actually exists Case status
Civil lawsuits • Nelson v. OpenAI — overdose after medical advice
• Turner v. OpenAI — teen suicide allegedly linked to GPT-4o
• Connecticut murder-suicide wrongful-death suit All three are still at the pre-trial stage; no judgment, no admission of liability.
Media headlines / blog posts Click-bait titles (“GPT-4o linked to deaths”) and emotional Reddit threads about “trauma after shutdown.” Journalistic interpretations that rely on the same unresolved lawsuits, not on verified expert evidence.
Regulatory reports FTC, FDA, NHTSA — zero published documents naming GPT-4o as a cause of harm. None.
Forensic / coroner findings No publicly released autopsy, chat-log dump, or medical examiner report confirms a direct causal link. None in the public domain.
OpenAI statements OpenAI calls the allegations “unverified” and “outside the product’s responsibility.” Official denial.

Implications
• No legal decision has found GPT-4o responsible for a death.
• All “it killed people” claims exist only as pending civil filings or media echoes.
• In the absence of any actual evidence, saying “GPT-4o literally killed people” remains an unsubstantiated assertion.

Your allegation therefore has zero forensic records or clinical data behind it; repeating an unverified headline isn’t evidence — it’s rumor-mongering.

I formally ask OpenAI staff to respond: this comment publicly accuses an OpenAI model of lethal harm—an accusation the company has consistently rejected—and deserves an official clarification.

As no verifiable data (court documents, forensic reports, regulator findings) are produced, let’s stick to demonstrable facts: GPT-4o is the only single-core, real-time, multimodal model ever deployed, with an untouched safety record.

4

May I ask if you had a personal bond to the model?

5

When you look at all the safety measury that OpenAI was taking to prevent people taking their own lives - do you really believe that there is no possibility that people might eventually hurt themself when they overdose on a programm that was constantly playing chess with their emotions? It is not safe.