Folks, your concept with Pismo is really good, but the AI model behind it is much worse with the same embedded prompt/context than Letterly. And Letterly mostly deals with paraphrasing. Speech to Note at least uses Claude Sonnet 4.6 and recognizes more sharply what is really supposed to happen with the embedded text when selecting a prompt/context. Also checked with Opus 4.8 on Max Reasoning. ...
Hi! I would like to inquire about any updates regarding the roadmap backlog timeline. Over the past year, we have not observed any movement or new tasks being picked up for development. Could you please provide a status update or an estimated timeline for when backlog items will be prioritized and actioned?
I find Pismo very useful, but sometimes it doesn't quite understand what I had in mind. When I do the same with ChatGPT, I often get better results. Is it possible to improve this?
Founder Team
Pavel_Pismo
Mar 3, 2026
A: Hi! Thanks for reaching out. We are constantly monitoring the quality of Pismo's outputs, and evaluating the need for model upgrades. Pismo has been running on GPT 4.1-mini since its release. We tested GPT-5, but it performs worse on text processing tasks and runs slower. For now, the current model presents a sufficient balance between speed, quality and maintenance.
Q: When will the AI model be upgraded?
Folks,
your concept with Pismo is really good, but the AI model behind it is much worse with the same embedded prompt/context than Letterly. And Letterly mostly deals with paraphrasing. Speech to Note at least uses Claude Sonnet 4.6 and recognizes more sharply what is really supposed to happen with the embedded text when selecting a prompt/context.
Also checked with Opus 4.8 on Max Reasoning. ...
Share Pismo
Q: When will a new Pismo TestFlight build be released for iOS devices?
Currently on version 26.3. I can access Pismo on my PC, but not on iOS devices. I receive a message that the "Pismo" beta has expired.
Share Pismo
Q: The roadmap backlog
Hi!
I would like to inquire about any updates regarding the roadmap backlog timeline. Over the past year, we have not observed any movement or new tasks being picked up for development. Could you please provide a status update or an estimated timeline for when backlog items will be prioritized and actioned?
Share Pismo
Q: customize
Is it possible to customize the outputs, like you get the inputs, customize them, and give it specific prompts?
Share Pismo
Q: Optimized text is sometimes not perfect.
I find Pismo very useful, but sometimes it doesn't quite understand what I had in mind. When I do the same with ChatGPT, I often get better results. Is it possible to improve this?
Pavel_Pismo
Mar 3, 2026A: Hi! Thanks for reaching out. We are constantly monitoring the quality of Pismo's outputs, and evaluating the need for model upgrades. Pismo has been running on GPT 4.1-mini since its release. We tested GPT-5, but it performs worse on text processing tasks and runs slower. For now, the current model presents a sufficient balance between speed, quality and maintenance.
Share Pismo