Gemini's New Argon Image Model Keeps Generating Distorted 'Little People'
RecognitionHefty · reddit · 2026-10-15
A user reports that Google's new Argon image model consistently generates figures with distorted, shrunken heads—unlike the previously reliable Pro 3.1. Asked to analyze the prompt, Argon admitted nothing suggested odd head proportions yet still produced the distorted output; Flash 3.8's editing is also criticized as poor. The user, previously a heavy Gemini user for imagery, is frustrated by degraded quality and rapid quota burn.
More from Models
- Meta's Muse Wonder (Max) lands at #12 in Agent Arena at 2.1x the cost of its predecessor — arena · 2026-10-15
- Scale CEO spotlights Muse Wonder: untrained for 3D yet beats Gemini 4 in physics visualization tests — alexandr_wang · 2026-10-15
- Cohere tops WMT26 translation benchmark: small agentic model scores 84.36, beating Mistral and DeepL — cohere · 2026-10-15
- Anthropic has never launched a Sonnet/Opus-tier model on a Wednesday — ChrisGPT · 2026-10-15
- 8k context beats 480k: OpenAI Decisions API cuts costs 84% at 98.25% F1 — OpenAIDevs · 2026-10-15
- Gemini 4 Argon hit with HTTP 524 timeouts outside Antigravity; Google payment also glitching — cedric_chee · 2026-10-15