Simon Willison’s Weblog

GPT-5.4 mini and GPT-5.4 nano, which can describe 76,000 photos for $52

17th March 2026

OpenAI today: Introducing GPT‑5.4 mini and nano. These models join GPT-5.4 which was released two weeks ago.

OpenAI’s self-reported benchmarks show the new 5.4-nano out-performing their previous GPT-5 mini model when run at maximum reasoning effort. The new mini is also 2x faster than the previous mini.

Here’s how the pricing looks—gpt-5.4-nano is notably even cheaper than Google’s Gemini 3.1 Flash-Lite:

I used GPT-5.4 nano to generate a description of this photo I took at the John M. Mossman Lock Collection:

llm -m gpt-5.4-nano -a IMG_2324.jpeg 'describe image'

Here’s the output:

The image shows the interior of a museum gallery with a long display wall. White-painted brick walls are covered with many framed portraits arranged in neat rows. Below the portraits, there are multiple glass display cases with dark wooden frames and glass tops/fronts, containing various old historical objects and equipment. The room has a polished wooden floor, hanging ceiling light fixtures/cords, and a few visible pipes near the top of the wall. In the foreground, glass cases run along the length of the room, reflecting items from other sections of the gallery.

That took 2,751 input tokens and 112 output tokens, at a cost of 0.069 cents (less than a tenth of a cent). That means describing every single photo in my 76,000 photo collection would cost around $52.44.

I released llm 0.29 with support for the new models.

Then I had OpenAI Codex loop through all five reasoning effort levels and all three models and produce this combined SVG grid of pelicans riding bicycles (generation transcripts here). I do like the gpt-5.4 xhigh one the best, it has a good bicycle (with nice spokes) and the pelican has a fish in its beak!