OpenAI: Our new model GPT-5-Codex-Mini – a more cost-efficient GPT-5-Codex

Re:OpenAI: Our new model GPT-5-Codex-Mini – a more cost-effi... by simonw on Saturday November 08, 2025 @10:42PM (#45862709)

I managed to get GPT-5-Codex-Mini to draw me a pelican. It's not a very good one! https://static.simonwillison.net/static/2025/codex-hacking-m...

For comparison, here's GPT-5-Codex (not mini) https://static.simonwillison.net/static/2025/codex-hacking-d... and full GPT-5: https://static.simonwillison.net/static/2025/codex-hacking-g...

I had quite a fun time getting those pelicans though... since GPT-5 Codex Mini isn't officially available via API yet I instead had OpenAI's Codex CLI tool extend itself (in Rust) to add a "codex prompt ..." tool which uses their existing custom auth scheme and backend API, then used that to generate the pelicans. Full details here: https://simonwillison.net/2025/Nov/9/gpt-5-codex-mini/

[ Parent | on HN ]

Re:OpenAI: Our new model GPT-5-Codex-Mini – a more cost-effi... by aurareturn on Sunday November 09, 2025 @12:48AM (#45863255)

Just curious, do you think LLM makers are deliberately adding training data for your pelican test by now?

[ Parent | on HN ]

Re:OpenAI: Our new model GPT-5-Codex-Mini – a more cost-effi... by simonw on Sunday November 09, 2025 @01:14AM (#45863366)

I'll know if they do, because I'll notice that a new model is suspiciously great at drawing pelicans riding bicycles while still sucking at drawing other things.

[ Parent | on HN ]

Re:OpenAI: Our new model GPT-5-Codex-Mini – a more cost-effi... by RestartKernel on Sunday November 09, 2025 @03:01AM (#45863790)

Can't believe I'm saying this, but GPT-5's pelican is its most impressive improvement over 4o I've seen. I wonder if Codex' MoE does not contain any expert fit for this task due to its fine-tuning on code.

[ Parent | on HN ]

Re:OpenAI: Our new model GPT-5-Codex-Mini – a more cost-effi... by wahnfrieden on Sunday November 09, 2025 @03:02AM (#45863792)

Make sure you check GPT-5.1’s from a couple days ago

[ Parent | on HN ]

Re:OpenAI: Our new model GPT-5-Codex-Mini – a more cost-effi... by karolcodes on Tuesday November 11, 2025 @12:20AM (#45884375)

wait it's not yet out, isn't it?

[ Parent | on HN ]

Re:OpenAI: Our new model GPT-5-Codex-Mini – a more cost-effi... by wahnfrieden on Wednesday November 12, 2025 @02:39AM (#45897341)

It's being tested already on huggingface under another name

[ Parent | on HN ]