Interesting that both "none" and "low" reasoning levels appeared not to spend any reasoning tokens at all, and in fact low ended up using less tokens and costing less than none here.

"medium" did use 496 reasoning tokens, but that's apparently not enough here to return a visible reasoning trace.

OpenAI's current frontier model lineup is gpt-6-astra, gpt-6.1-sol, gpt-6.1-terra, gpt-6-luna, so here's a comparison grid of all of their pelicans at different reasoning levels: https://static.simonwillison.net/static/2026/openai-pelicans...

▲

AISnakeOil 5 hours ago | parent | next [-]

What's the point of using it without reasoning? Isn't that what Luna's for?

	▲	simonw 5 hours ago \| parent \| next [-]
		It's neat that we can now compare the Terra without reasoning pelican (cost 0.64 cents) with the equivalently priced Luna with reasoning pelicans (Luna xhigh is 0.40 cents, Luna max is 0.91 cents). I think the Terra low one compares favorably to those, though I'd suggest also comparing them on tasks more representative of real work than generating pelican SVGs!
	▲	notatoad 5 hours ago \| parent \| prev [-]
		a bigger model still has more built-in knowledge to draw from, even without reasoning.

▲

shonenknifefan1 4 hours ago | parent | prev | next [-]

I appreciate that most of these pelicans are not holding the handlebars, i think that's more realistic than the alternative

▲

pietz 4 hours ago | parent | prev | next [-]

I can literally see what the training data sample looked like that they included, so it performs well on this benchmark.

▲

drewbitt 5 hours ago | parent | prev | next [-]

12 seconds for a pelican on low. Pretty good.

▲

kubb 5 hours ago | parent | prev | next [-]

Do you have an agent scanning the front page waiting to post the pelicans?

	▲	simonw 4 hours ago \| parent [-]
		I should do but no, it's just a bad HN habit.

▲

bigcat12345678 5 hours ago | parent | prev [-]

Seems Gemini is way ahead in pelican drawing

	▲	thimabi 5 hours ago \| parent [-]
		Apparently Google made it a point to train their models to output SVG, among other things, so that helps