Every ERNIE post, filterable by category, sortable by date, searchable by keyword.
The ERNIE family splits across two platforms. Qianfan serves ERNIE 5.0 text and multimodal from China. fal serves ERNIE-Image globally. Here is the exact routing call.
Flux 2 Pro and GPT Image 2 are frontier general purpose models. ERNIE-Image is a specialist. On dense Chinese signage, bilingual menus, and Japanese posters, the specialist wins by a visible margin.
English glyphs on ERNIE-Image posters have a kerning problem. Here is why the model behaves that way, what prompt patterns fix it, and when to switch to GPT Image 2 instead.
ERNIE 5.0 landed on January 22, 2026 with a 2.4T MoE architecture and native omni-modal IO. Here is what changed versus 4.5 Turbo, and when the older line still earns its keep.
Two headline numbers tell you where ERNIE 5.0 sits: LMArena 1460 at rank 8, AA Intelligence Index 29. Here is what both actually mean for your product.
One endpoint charges $0.03 per megapixel and runs 50 diffusion steps. The other charges $0.01 and runs 8. Here is where the 3x price delta earns its money and where it quietly burns your budget.