In my experience, it is claude-code paired with deepseek-v4. For penny-pinchers like me, I can have long coding sessions with it with no anxiety about the cost. Also, prompting it to what you want and verifying the outputs is more important than the quality of the model. So, I am better off with a cheaper model and taking the responsibility for prompting it and verifying the results.
I've gone through ~600m tokens in Xiaomi Mimo though Claude, and it's been the most effective use of an agent I've had yet. It's very capable, but generally not ambitious, picking simple but effective solutions to most problems I give it.
Going to write something longer about the experience when I get to a billion tokens.
I use it through my opencode go subscription and it's exactly how you described. Very pragmatic and not too ambitious. It's similar to Kimi 2.5/6 in that regard.
Although I have little interest in agentic coding, when I do use it, I have found Kimi K2.6 to give Opus-quality output, and have switched entirely to it for pretty much everything.
I've used Opus extensively and tried K2.6 on a few projects, and the gap is huge. K2.6 is nowhere near the performance of Opus. That's fine because it's also far cheaper, but public benchmarks line up with my own personal experience that they aren't comparable in terms of intelligence.
(that is, different places on the Pareto efficiency graph)
No two uses are alike, I suppose. For me, whatever difference is a wash. However, I probably tend to shy away from throwing high-complexity/long-horizon tasks at the model.
I'd generally agree about Deepseek being as good as Sonnet - but I have extreme trouble with prompt compliance with V4 Pro in a way that I've never had with Sonnet. I'll tell it "find the bug, but don't fix it" or "please use this tool I just developed" and it'll ignore me a high fraction of the time.
It's bad enough that I'm working on guardrails at the harness level because prompting appears to be useless.
I have Opus make a fairly detailed plan, then Deepseek implements, and GPT reviews. With that setup, I have zero issues, probably because what you mention is handled (the plan keeps it on track and the reviewer catches any issues).
Now that you mention it, though, I have seen it do a few things that weren't in the plan. The reviewer caught them, though, so they didn't cause a problem, and it's so cheap that overall it's a massive improvement.
It's the only model where an explicit instruction at the end of my message is sometimes ignored. This doesn't happen with any of the gpts, kimis, glms, qwen, etc. Just a deepseek problem.
I have also noticed this with Sonnet, funnily enough - it's not as strong, but it's still there. But yeah, I haven't seen this with any other model so far (although I mostly use the stronger ones - maybe it's a function of intelligence?).
Cursor with Composer 2.5 seems to be competitive with frontier models (Opus and GPT-5.5) for a significant price discount. Benchmarks are gamed, as always, but $0.55/task vs $11.02 a task definitely indicates that there's some cost advantage.
