Competition is good

It really is good. I forgot to mention that within that said sub agent, it also went into exploring top e-commerce websites (Zalaondo, Temu, Amazon, eBay) for exploring prevailing industry UX best practices and taking screenshots of their product and category pages with its own written chrome driver that I talked about and then went onto prototyping a new website in a temporary directory and then taking hundreds of screenshots to analyse what would be the best column density one each medium for each language.

And that all is 0.07 cents all included.

What harness do you use with it? Are you using v4.1flash via open router ?

I am using DeepSeek Harness[0] (switched from OpenCode) and I am using DeepSeek directly via the API. The speed is insane. Like 200 tokens/second is the norm but I have seen much higher too at times.

PS: I do not know why but opencode pushes CPU usage to very high which has NOT happened with DeepSeek harness even once.

[0]. https://github.com/deepseek-ai/deepseek-harness

That's incredible, might just switch myself. Thank you!!

> and I am using DeepSeek directly via the API.

Do you know if they retain your prompts or use it for training?

They train. Deepseek doesn't have a Zero Data Retention policy, but for many personal tasks I'm (not parent) ok with them having this data if it means better, cheaper models for everyone on the future.

Different story with $DAYJOB where ZDR matters, but the beauty of open models is that you can find a provider hosting them in your preferred terms.

Also include that all of this comes with full reasoning traces, so if something goes wrong, you know exactly what assumption it started from.

Yes exactly. Reading this "thinking" traces is a great tool.

Compared to Opus 5, and others, I also found DeepSeek 4.1 Max to be really good and cheap. I am testing right now with Opus 5.5 and I feel it way cheaper than Opus 5!

Same here! DeepSeek v4.1 Flash has been my moment of "does everything I need, cheaply. Please now focus all R+D on making this efficient enough to run off a laptop"

Assuming you are not using DeepSeek with Claude code so what/how are you using it?

I honestly don’t think deepseek is as good inherently as we treat it… but it thinks so much so fast it gets itself there. I’ve been super impressed.

If you can give it a measurable objective, load up $5 or so and just leave it for a few hours it usually gets there.

What are you working on? Im always a bit surprised by folks that seem ok with non frontier models - the quality is just not there. I've found most code produced by even luna / sonnet tier models to be significantly worse quality. It seems to me they can't handle any mild complexity at all. Are you just prompting very explicitly and detailed?

Yes, I tell exactly how I want it done and what files to modify, in those conditions sometimes a model that does not try to read between the lines works better.

I’d not put Luna and Deepseek in the same tier as Sonnet, they were clearly ahead last time I checked (though I might be outdated and that’s on my personal use case).

I’ve been working with frontier models for the last year on a large real-world project with multiple apps. Eventually they started becoming unreliable. Simple UI issues, usage limits and overengineering became constant problems. I had to come up with a robust process: task analysis, user intent analysis, implementation and testing. All of these steps loop when needed with 15-20 steps per task in total. This finally got the models to do a good job but I started hitting usage limits. Then I switched to DeepSeek and haven’t noticed any drop in quality. It’s fast at around 300 TPS and cheap. I don’t miss frontier models anymore! Come up with a good process and you might not need them either.

How much do you get out of these models for same amount (say $20 sub for claude/codex/zai-glm) - the equivalent amount of tokens? Also is deepseek as good as glm5.3? (if we don't compare it with frontier models from USA)

Well, to be honest it’s all based on my subjective experience. I started benchmarking models on 10 of my real tasks ranging from easy to hard but even with the same model the completion time sometimes varied by as much as 50% between runs. That made me realize you’d probably need hundreds of tasks to get any meaningful numbers. Otherwise there’s just too much variance. So I can’t give you solid benchmarks without spending weeks testing everything properly. What I can say is that I actually use most of these models regularly for different things, so I’ve developed a general feel for them. DS is my main model for work, Claude for hobby projects and local AI, Fable mostly for design, Codex for various other tasks and sometimes Grok for health-related stuff. Cost-wise, a $200 subscription wasn’t enough to get through 20-30 tasks while $30 of DeepSeek was, and that was with Sol, not even Astra. Fable is expensive too and I haven’t found it particularly strong at coding. Opus 5 was horrible in my experience, though maybe 5.5 will be better since I’m testing it now. I also ran GLM 5.3 Flash locally for a while but it was pretty slow and its code was usually worse than both DS4.1 and Qwen Flash Next. Things are moving so quickly that it’s honestly hard to keep up, so take all of this as my general experience from actually using these models rather than a proper benchmark.

> Cost-wise, a $200 subscription wasn’t enough to get through 20-30 tasks while $30 of DeepSeek was

Thank you. This alone helps! I think I should jump into it once and try it all out while I still have GLM access at old prices as a backup.

Np, good luck! PS I’ve been using Sol 6 and Opus 5.5 today and they’ve been great so far. The new Opus is smart, fast and cheap. It makes mistakes but after every task I ask Sol to review it and it catches most of them. I need to test the models properly but it looks promising so far! Hope they won’t nerf them again. I think DS may still be cheaper though (and it’s 300 TPS via the API which is sick).

This is so true, I use frontier models at work and the variance in the day to day experience can be dramatic.

I use Deepseek for some home stuff, and it's as you describe - fast, consistent. Not brainiac level, but when you drive it well it just tends to execute reliably. I have seen it fail consistently on very hard stuff, but I'm certainly ok with having to special case occasionally. Especially if those occasions keep getting less and less numerous as the models improve.

Now I need to convince my company to allow us to use a Chinese model ... I don't think I'll have much luck.

If you're producing slop, the quality of the model is irrelevant as long as it compiles. I frequently catch Opus/Sol making silly errors and over-engineering solutions while small ones like Luna struggle with complex tasks.

Deepseek on high reminds me of Opus 4.x which was quite good for a lot of tasks.

Incredibly cheap. What harness/orchestrator are you using?

OP replied to a comment of mine, they use Deepseek harness:

https://github.com/deepseek-ai/deepseek-harness

Deepseek is cheap and fast, but it couldn't follow the basic directions I gave every other model (GLM, Qwen, GPT, Claude) to use red/green TDD. Put me off of using DS.

Why did it need to generate its own driver when there is a Playwright MCP available and works right out of the box? Am I missing something?

Hey. I am on a GLM Coding plan subscription (old price; their base coding plan) right now and share the key (this will go away soon).

I was thinking of going with a subscription of Claude or Codex. The reason (at least that's what I am assuming): with OpenRouter or any PAYG per token setup there will be the anxiety of using up all the tokens in days or maybe 1-2 weeks instead of a month (say I set myself a budget of 15-20 USD per month, average equivalent of a usual subscription price).

Now I don't really want the top-notch models for the coding work I do.

So how much worth of "work/tokens" will I reasonably get for ≈$20 USD if I use it a lot? How much does that equal to - or is equivalent to, say in the world of subscription based Claude, Codex, or even GLM (with their 5-hour and all those cooldowns/limits)?

I am looking for a mental model/framework to visualise this. Can you (or anyone else reading this) please point me to a source where I can get some idea about this? I know I can just add $5 on OpenRouter and try to test. But I don't really know what/how to test these spends. I also want to understand how all this works. (I am new to agentic/llm world/coding, 2-3 months, after a career break of ~3 years, that too after working for more than a decade. I know, not at all good timing!)

A good resource is this: https://artificialanalysis.ai/?cost=intelligence-vs-cost-per...

Go to the "Cost" -> "Intelligence Index vs. Cost per Intelligence Index Task" That diagram maps their "Intelligence" score to "cost per task" and I think this gives a good basis on deciding with which model you want to go. Then you can either get an API token from that models provider directly or use openrouter and set openrouter to the model/providers of your choice.

You can also see on openrouter itself the details for each model like prices and what providers are offering it at what price.

Finally you can compare models details using openrouters compare feature like this:

https://openrouter.ai/compare/anthropic/claude-opus-5.5/open...

This is very helpful. Thank you!

what harness do you use? I do want to try out deepseek but don't know what the equivalent of codex/claude code is here

> which failed to drive chrome so it wrote its own chrome driver protocol server

This is one of the things I hate the most. Super complicated workarounds which take loads of time (and sometimes money) for even the simplest problems. Human would pause and ask. I would blame harness, not the model though.

Insane!