Tool Ren'Py Ren'py realtime translator Mod[v0.47.b]

[NaveenKrishnanC]

View attachment 5552548 View attachment 5552550
1. Sorry, where can I find those code?
2.

You must be registered to see the links

is a must?
3. I am gonna use AI studio API key and Gemini flash 2.5. Should I input "Gemini flash 2.5" in model_name?

1-The guide in the first post is out of date now.
You can search "openrouter" in realtimetrans.rpy to locate the lines.
2-No.
But i did not test other providers，you may need to test yourself.
3-I think gemini is not supporting openai type requests ，so it probably not works.


I've been very busy at work this week, with increased maintenance due to the Christmas holidays weather.
Therefore, updates won't be frequent.
I'll try to release an update this weekend and update the guide to the latest version.

 

[NaveenKrishnanC]

View attachment 5552548 View attachment 5552550
1. Sorry, where can I find those code?
2.

You must be registered to see the links

is a must?
3. I am gonna use AI studio API key and Gemini flash 2.5. Should I input "Gemini flash 2.5" in model_name?

Tested AI studio API key and Gemini flash 2.5.
1-In realtimetranslrpy,you should find

("api_keys", ["Your-API-KEY1", "Your-API-KEY2"]),#api,model,keys can not be set within the game ,you need to edit this file to change them.
 ("model_name", "meta-llama/llama-3.3-8b-instruct:free"),
 ("base_url", "https://openrouter.ai/api/v1/chat/completions"),

2-replace code with

("api_keys", ["my GEMINI api"]),#api,model,keys can not be set within the game ,you need to edit this file to change them.
 ("model_name", "gemini-2.5-flash"),
 ("base_url", "https://generativelanguage.googleapis.com/v1beta/openai/v1/chat/completions"),

3-Set the time_interval properly to meet google's rate limit.
In my test using free api,i encountered error 429 RESOURCE_EXHAUSTED very soon.
You might need to pay for increasing rate limit .

 

[NaveenKrishnanC]

Tested AI studio API key and Gemini flash 2.5.
1-In realtimetranslrpy,you should find

("api_keys", ["Your-API-KEY1", "Your-API-KEY2"]),#api,model,keys can not be set within the game ,you need to edit this file to change them.
("model_name", "meta-llama/llama-3.3-8b-instruct:free"),
("base_url", "https://openrouter.ai/api/v1/chat/completions"),

2-replace code with

("api_keys", ["my GEMINI api"]),#api,model,keys can not be set within the game ,you need to edit this file to change them.
("model_name", "gemini-2.5-flash"),
("base_url", "https://generativelanguage.googleapis.com/v1beta/openai/v1/chat/completions"),

3-Set the time_interval properly to meet google's rate limit.
In my test using free api,i encountered error 429 RESOURCE_EXHAUSTED very soon.
You might need to pay for increasing rate limit .

And for guys who want to use LLM,
Check your providers' docs.
ONLY IF your api provider supports open-ai type requests,LLM is possible.

Check the example python codes.
base_url should be set to end_points url ,usually it is
url,eg. google provides "

You must be registered to see the links

"
+"v1/chat/completions"
api_key and model_name should be same as example python code.

 

[NaveenKrishnanC]

v0.44.a
update:27-12-2025
updated file:realtimetrans.rpy
1-Add prescan for part of the menu texts .
2-Prescan is now working in sub process,so it should not block game in the beginning now.
3-Add variables apply for prescan texts,
if there are variables in prescan texts,
it will be replaced with its value and added to translation_cache.

 

[NaveenKrishnanC]

Guide is updated.

 

[NaveenKrishnanC]

v0.45.a
update:30-12-2025
updated file:realtimetrans.rpy
Add another freellm provider: api.deepai.org.
This provider is found in webscout

You must be registered to see the links

In my test ,this provider is stable and could response even more frequent requests than
oi-vscode-server-0501.onrender.com,which is often blocked by cloudflare for too frequent requests.

"deepseek-v3.2",
"gemini-2.5-flash-lite",
"qwen3-30b-a3b",
"gpt-5-nano",
"gpt-oss-120b",
"llama-4-scout",
"llama-3.3-70b-instruct",
"gemini-2.5-pro",

The models in the list will be randomly selected if this provider is chosen randomly.

 

[Luci3per]

Hi,

Can anyone help me with this, please?

All 4 files were newly downloaded and copied to the game folder. None were modified. It works for other games. (I think maybe not for 2 of them). Log and traceback attached.

Thank you.

Translate by DeepL.

 

[NaveenKrishnanC]

Hi,

Can anyone help me with this, please?

All 4 files were newly downloaded and copied to the game folder. None were modified. It works for other games. (I think maybe not for 2 of them). Log and traceback attached.

Thank you.

Translate by DeepL.

GAME NAME,please.

 

[Luci3per]

GAME NAME,please.

Crimson Prison
https://f95zone.to/threads/crimson-prison-v0-1-5-0-dajimess.264356/

 

[NaveenKrishnanC]

Crimson Prison
https://f95zone.to/threads/crimson-prison-v0-1-5-0-dajimess.264356/

v0.46.a
update:03-01-2026
updated file:realtimetrans.rpy
1-Fixed a few bugs in Renpy7.
2-Fixed one bug encoding non-ASCII characters.
Now the mod should work well other than English using google/freellm/LLM.
Edge is still only for english now.
3-Fixed a bug in prescan ,which will crash the game in android.

 

[Luci3per]

Yes, this upgrade works, and it also started working with the second game that I remembered not working: Chronoromancer.

Thank you very much for your help.

 

[denadm]

Is there a function to save translated text, or does it have to be requested online each time?
For example, to translate something in order to share a novel with the translation.

 

[trans1]

The translated texts are stored in the translation_cache.json file.

 

[NaveenKrishnanC]

v0.47.a
update:07-01-2026
updated file:realtimetrans.rpy
1-Fixed bracket_pattern bug, now [] variables should be protected properly.
2-oi-vscode-server-0501.onrender.com is not available now as it updated auth.
Freellm is always using deepai now.
3-prescan_texts is not saved persistently now ,this may crash your saves.

 

[NaveenKrishnanC]

v0.47.b
update:08-01-2026
updated file:realtimetrans.rpy
1-Appended lines will be history dialogues if there are enough dialogues, rather than lines translated before.
2-Fixed a bug freellm deepai not working in Renpy7 which is solved in v0.45.a butnot applied since v0.46