GPT-4o (OpenAI)

A dated reading of what is claimed, reported, and independently verified in the current evidence.

Current reading

Readiness band and full integer triple

AI-assisted assembly · derived results

Readiness band

Mature

Primary summary from verified readiness

Evidence strength Growing

Decision

Human editorial judgment · 2026-07-23

Track; not yet

Why: GPT-4o was voluntarily retired by OpenAI (2026-02-13) after collapsing to ~0.1% usage share and being cited as OpenAI's worst model for sycophancy amid self-harm-related lawsuits. It is historically important as the direct architectural predecessor of today's natively multimodal flagships, but the model itself is end-of-life and not a live adoption candidate.
Next: No direct adoption action for GPT-4o itself — it is retired from ChatGPT and being phased out at the API level. Track the architectural pattern it pioneered (single model, all modalities) as already reflected in current flagship choices (Claude Opus 4.x Vision, GPT-5.x, Gemini 3.x) rather than integrating this specific model.

Constraints

No named blocker is present in the current public projection.

Evidence summary

AI-assisted assembly

Counts and dates only. Raw signals, private excerpts, trust records, and internal corpus material are not published here.

Publication record

Initial public reading