this post was submitted on 03 Oct 2026
479 points (98.2% liked)
Technology
88559 readers
3284 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 3 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
AI works. But China already won the race. These prices are absolute peak and from now one will only be falling and that's not what openai and anthropic expected.
I can just use glm 5.3or depseek 4.1 and 20$ will go the distance.
Even running Qwen 3.8 locally can do 90% of what I want.
Yeah, I use 3.8 27B and it’s like Sonnet 5 for a lot of tasks. If they’re keeping up with Sonnet, 5.5 is similar to Fable, and it would be pretty nuts if Qwen 4 27B kept pace.
I have been using the Strata backend with Next-Flash. It has been putting together a Vampire Survivors golden egg manager for me, and I do the QA testing. I am looking forward to Qwen4-Flash, it should open up more possibilities and take off some rough edges. Maybe I can finally do a recompilation of Stars!, and other obscure games?
I literally just learned about Strata. Between that and FreeToken the LLM serving side has become so much better.
How does one get to use deepseek? Alibaba Cloud is always out if subscriptions.
I use deepseek via openrouter and deepseek direct. many american providers offer deepseek.
also from deepseek directly - works great with opencode / openchamber
Isn’t that kinda expensive as fck? I tend to use all the weekly usage of ChatGPT 20x within 1-2 days, and I already work with different gpt models, with different reasoning levels. So I am not sure if I can afford OpenRouter to be honest.
Not expensive with those models. Glm and deepseek. Problem with chatgpt they train you your data, basically siphon what you have and someone else could build your product faster than you. There is a recent case theybstole some scientific breakthrough.
Openrouter has zero data retention and zero training. Which is why I use it.
Chatgpt can't be used by companies
Will give it a go. Thanks for sharing your experiences.