Show HN: I built a fast, accurate translation API for 1/30th the price of Google (translateapi.web.app)
Hey HN,
I built an alternative to Google translate that's fast, accurate, and 1/30th the price. 20+ supported languages, not billed until 100k tokens, and already used thousands of times with real language learning users.
The landing page has a demo for you to test performance yourself for free with no card required. (Also landing page says 1/13 the price because had to account for all 3 big names, with varying prices each).
I built this as a workaround for my language learning startup, and it was perfect at a fraction of the cost. Hopefully it can help other devs in accessibility and language learning apps.
Thanks,
14 comments
[ 2.8 ms ] story [ 42.7 ms ] threadMy goal is to keep fundamental utilities and accessibility tools like translation services in the hands of small creators rather than some mega corporation, so I'm open to any improvement that can make that a reality.
Nic
Also read your twitter bio: I wont repeat here for privacy but would enjoy seeing stories about those things posted on HN
Thanks! Just for clarity, are you talking about the work experience, living situation, or startup ventures?
A few questions/suggestions:
- It's not clear what the character limit would be if one was to sign-up. - What does "context-driven translations for higher accuracy" mean? - 35 characters are quite short for testing the translation of a complete sentence. Is there a reason for using such a short character length?
The character limit for an api call is 3000. I put it in the docs but could probably make it more obvious.
The context driven for higher accuracy is referencing what's going on under the hood, but on second thought it doesn't make much sense to put that there if I'm not going in depth about the under the hood stuff. I'll take that out.
35 character is quite short, you're right. I'll double it. The reason it was so short was arbitrary. I made it 100 now.
If you've got anything else send it over !
Running an opensource AI model to perform translation may help and work in some cases but to be honest this has also a lot of limitations. The first being processing time and the second languages support. I am not even considering accuracy.
Did you try the demo on the landing page and notice considerable latency/accuracy slip ups?
How does it work? Is it essentially a fine-tuned LLM? If so, which base model?