Talkie
Fewer than 5 people recommend it so far, so there's no written verdict. Here is what commenters said.
3recommend
0warn against
Apr 27, 2026first mention
Sep 3, 2026latest mention
Why people recommend it
“Talkie[1] presents an interesting case for it being at least possible”
“You can try it here: https://talkie-lm.com/chat”
Show 2 more
“Parakeet and incredibly fast and made for devs”
Caveats and dissent
No one in these comments warned against it.
From the makers not counted in the ranking
“built my own but with a bit more emphasis on speed”
Threads where it came up
- Talkie: a 13B vintage language model from 1930 3 mentions
- Show HN: Free Alternative to Wispr Flow, Superwhisper, and Monologue 2 mentions
- K2 Horizon: A connected fleet of six open models 1 mention
- What happens when an LLM never sees material beyond fifth grade? 1 mention
- Position: LLMs Can't Jump 1 mention
- GPT-5.6 used a prompt to close a 30-year gap in convex optimization 1 mention
Also written as talkie. Data: talkie.json. Quotes are verbatim; the verdict and theme labels are written by a model from these comments, and every clause links to its sources.
All comments 11
“Talkie[1] presents an interesting case for it being at least possible”
“A similar project (LLM trained only on vintage material)”
“They did that with "Talkie", a model trained on 1930 and before.”
“Even really rubbish models, e.g. talkie”
“A model pretrained on texts with 1930 data cutoff can solve a few programming problems”
“This LLM trained only and entirely on pre-1930s texts was able to code Python programs when given only a short example”
“You can try it here: https://talkie-lm.com/chat”
“I'm trying to convince TALKIE about computers.”
“I love this.”
“Talkie isn't entirely pure either though”
“Parakeet and incredibly fast and made for devs”