The pelican is excellent for a 16.8GB quantized local model:

simonw • today at 4:46 PM • 4 replies • view on HN

The pelican is excellent for a 16.8GB quantized local model: https://simonwillison.net/2026/Apr/22/qwen36-27b/

I ran it on an M5 Pro with 128GB of RAM, but it only needs ~20GB of that. I expect it will run OK on a 32GB machine.

Performance numbers:

  Reading: 20 tokens, 0.4s, 54.32 tokens/s
  Generation: 4,444 tokens, 2min 53s, 25.57 tokens/s

I like it better than the pelican I got from Opus 4.7 the other day: https://simonwillison.net/2026/Apr/16/qwen-beats-opus/

throwaw12 • today at 4:47 PM

I feel like this time it is indeed in the training set, because it is too good to be true.

Can you run your other tests and see the difference?

➕ show 4 replies

ahoog42 • today at 5:53 PM

at what point do model providers optimize for the "pelican riding a bicycle" test so they place well on Simon's influential benchmark? :-)

➕ show 2 replies

halJordan • today at 8:08 PM

These are the stupidest things to cleave to.

echelon • today at 7:42 PM

Metrics and toy examples can be gamed. Rather than these silly examples, how does it feel?

Can you replace Claude Code Opus or Codex with this?

Does it feel >80% as good on "real world" tasks you do on a day to day basis.

alt Hacker News