The speech limitation sounds difficult.

As of earlier this year, with the release of Mordenkainen Presents: Monsters of the Multiverse and the discontinuation of digital copies of Volo’s Guide to Monsters, kenku do not have speech restrictions or limitations.

Personally, I like the voice restriction, but I understand it can be hard to roleplay around.

One thing you can do is, instead of voicing everything the Kenku says, just give a summary of their intent. Something like, "In a variety of voices, my Kenku explains that he saw the corrupt politician accepting a bribe from a member of the Thieve's Guild".

Something else I did was have two distinct voices that my Kenku character would use depending on the situation... the idea being that these voices just come from two different people he had spent enough time with to pick up a lot of phrases from. So I did a bad cockney accent whenever talking about anything crime-related, because the character was a former criminal, and also a scholarly voice when explaining anything related to arcana or history, since the character's backstory was that he did work at a college copying textbooks and would listen in on a professor giving lessons. I didn't worry about making sure that I stuck to just phrases heard in-game or anything... as long as I presented information as though I was "in-character" as these other characters it still helped to convey the idea that my Kenku was impersonating someone else.

From there you can just use being a Kenku as an excuse to insert silly quotes or voices into the game. I remember for a brief period I actually had a sound board of Schwarzenegger one-liners loaded up on my phone that I could use to get a laugh out of other people at the table.

I made a Kenku that had a pretty limited pool of answers pre- written down. It was in his backstory, so he did only ever use the words spoken to him by his dead friend, so this list was all he had. Everything else I did with the imitation of sounds (instead of asking what does it cost /pay he made the sound of rolling coins etc.), but I often used "he imitates XY to say Z". It worked pretty well, the only problem I had was my List didn't contain ANY questions.

It worked pretty good.

For my kenku I would describe sounds to convey a message.  A cats hissing= danger .The sound of leaves rustling = stealth/hidden.  People's "names" would be them introducing themselves or a common phrase.

OMG, several years back we had a kenku in our party that was played by a genius.  Right from the get go he listened to other characters saying things and tried to mimic their voices and their words every time he wanted to say something.  He actually mimicked everyone else at the playing table jumping between voices and phrases that we had said to piece together the things he wanted to say.

Understandably he couldn't keep it up forever but after the first session he would just mimic a few of the more outrageous things we or the DM had said.  That was some of the most fun we ever had.

This is what I said in a far older thread about how I played my Kenku:

I'm playing a Kenku right now. Here's my advice based on my current experience:

 

- Make notes of anything you think might be useful or entertaining that's said during your campaign.

- Make a list of phrases that you think make sense and can cover specific scenarios.

 

For any other situation where you want to contribute just make something up. It's not beyond the realm of possibility that in the time prior to the start of the campaign (ie. the beginning of this leg of the character's life story) they haven't heard any countless number of voices, phrases, sentences, and sayings.

Don't limit yourself specifically to a small list of pre-concocted options or, even worse, only what you've heard since the start of the campaign as it risks being far less fun to play.

And if you want to, have fun with voices and accents. My fellow players seemed to enjoy my mimicking the voices and accents they were using for their characters especially when the things they were said were repeated at the most inopportune or inappropriate times such as when the guards were questioning us about a possible murder, rushing us for an answer to one particular question, and I say to the guard without hesitation "Keep your [expletive deleted] pants on" in our Life Cleric's voice, or as close to as I could manage.

Could it be possible that a Kenku reads many different phrases and listens to one specific voice for long enough to use the phrases read in books but in the targeted voice? Could you somehow use that to build lines for a Kenku to say? I am just curious as Kenku are one of the races I haven't looked into much yet but they seem pretty interesting. Actually, as I type this, I had this idea. I watched an anime recently where the MC can't understand the common tongue, but he can mostly understand the written language. So he uses magic to write the words he wants to say. Since Kenku's have unnaturally good memory (At least that's what i think i heard) They could memorise a large list of phrases and words and use like sending stones or some kind of written means to convey messages. Also if they imitated a dinging noise when they write something it could be like a notification noise so the party members know that the Kenku wants to say something. Outside the game mechanics would be similar while roleplaying. Just have the Kenku player use a whiteboard or something whenever they want to say something.

Anyone have any thoughts on this idea? I would appreciate some feedback on my suggestions and maybe people could fine tune it for their own campaigns and roleplaying.

I think since the whole curse isn't RAW anymore you can design it as you want. The whole ding and writing thing would work I think, but I think you shouldn't go so far as to use a whiteboard, that would slow down the game considerably and be ultimately annoying for most people. I's just say "he has written: 'Hi, my name is Bleach' ".

The whole talking with another voice in words you don't have specifically heard the person say but read another time is interesting to me, beacaus you could argue that all words are made out of syllables and once you'd have heard all of them you should be able to 'remix' them, but it stretches the whole speach restriction by quite a bit, so far as to say that you could just not use it and go with the MMoM version.

fair point thanks for the feedback.

I really appreciate the replies! It really feels great ot be a part of a community like this!