@deepseek_ai Fantastic news. I can't wait to try it. Things are moving so fast. I was just about to go back to OpenAI for the new Luna rates! Maybe i'll hang tight for a bit...
🚀 DeepSeek-V4-Flash Official API is now LIVE in public beta!
🔷 We’ve massively upgraded its Agent capabilities—benchmark scores are now far surpassing the V4-Pro-Preview. Check out the massive performance leap below! 👇
🔷 The official V4-Flash now natively supports the Responses API format and is fully adapted for Codex!
Check out the configuration details in our official API docs: https://t.co/smCwQZMeiq
🚀 DeepSeek-V4-Flash Official API is now LIVE in public beta!
🔷 We’ve massively upgraded its Agent capabilities—benchmark scores are now far surpassing the V4-Pro-Preview. Check out the massive performance leap below! 👇
🔷 The official V4-Flash now natively supports the Responses API format and is fully adapted for Codex!
Check out the configuration details in our official API docs: https://t.co/smCwQZMeiq
@Da7_Tech@ivanfioravanti maybe, but are they going to improve their coding plan inference speed and limits? Its pretty bad right now.. this Luna price change really makes me thing. Luna max is supposed to be as good as GLM 5.2 right?
@Da7_Tech Ive been on opencode for months now and haven't even attempted to try another one. My project is a personal OS that I started way before hermes was around and its so baked into opencode I can't leave! I steal everything I like from hermes, though. Still having tons of fun.
@Da7_Tech And I fully expect to cancel when they start billing full price for qwen 3.8. I hate these token plans that don't tell you how much usage you get. Its real bad with Alibaba. They don't even give you a clue!
@Da7_Tech I posted this a couple of days ago regarding usage on that plan:
I just finished my first week on the $18 Alibaba Token Plan using Qwen3.8 Max Preview exclusively. I got up to 98.9% of my weekly quota across 60 sessions all in opencode: 16.3M input 1.2M output 340.1M cached
@OctenAI Oh heck yeah! I just signed up. Let's go! I'm addicted to search tools and can't wait to add this one to my mix. High expectations for the guy who created Alibaba search though!!!
I just finished my first week on the $18 Alibaba Token Plan using Qwen3.8 Max Preview exclusively. I got up to 98.9% of my weekly quota across 60 sessions all in opencode:
16.3M input
1.2M output
340.1M cached
(saving for when they end the preview token rates so I can compare)
@OmedVibeCodes I don't trust a single one of these providers. The phrase "if it sounds too good to be true....." really rings here. I knew it wasn't going to last long. lol. They got $18 from me. That's enough.
@Da7_Tech Love this question. A balance of "smart enough" and how much usage I get. Models have gotten so good that my only project, which is a personal OS, doesn't need the best anymore. I want the best, tho! What about you?