AnecdoteArticle

Using DeepSeek-V4-Flash-0731's default reasoning level via OpenRouter produced a disappointing pelican image, while raising the reasoning effort to high yielded a much better result.

Recounts a hands-on test: the default reasoning level via OpenRouter gave a disappointing pelican image, but bumping reasoning level up to high produced something much better. ✦ AI generated

Author · Simon Willison's Weblog · 2026-07-31 · original ↗

Share card ↓Export .md Export .txt

I got a disappointing pelican from it using the default reasoning level via OpenRouter: But when I bumped reasoning level up to high I got something much better: llm -m openrouter/deepseek/deepseek-v4-flash-0731 -t pelican -o reasoning_effort high

Read full article ↗excerpt · fair-use quotation

▦ OpenRouter ✳ DeepSeek V4 Flash 0731

Around this claim

Mechanism · 1

Model size and reasoning effort are two separate scaling knobs whose performance curves overlap — a smaller model run at higher reasoning effort can sometimes match a larger model run at lower effort.Sebastian Raschka · Ahead of AI · conf 60%