Do they have an API where you can control the chat template or at least just put everything in the system prompt? This way you can control everything including the tool calling syntax. Even if you use the trained tool syntax, it allows you to control the tool system prompt which you may want to tweak. With DeepSeek this is all possible. An undocumented feature, great for harness builders. Anybody got info on Qwen regarding this?
Qwen3.6-Plus: Towards real world agents
171–180 of 235 posts
Re: Qwen3.6-Plus: Towards real world agents
#172I would love to hear from people using both (Claude Code OR Codex) AND (Qwen) and their experience with Qwen models, are they on par, or how far are they?
I switch between Claude Code (Opus/Sonnet) and Qwen (OpenCode, OpenClaw) multiple times throughout the day and Qwen 3.5 is really nice. I do also use KimiK2.5 and GLM5 pretty often too and I'm starting to get a sense that the agent tool is becoming a little more important than the model with these level of models. As long as tool calling and prompt quality is all configured correctly by the provider.
What are the reasons for switching? Personally I got into the habit of doing a bit of a round robin with Codex/Claude (CLI) and then DeepSeek and Qwen web chat. And Claude in web chat. I like to switch just to learn the differences, otherwise I'd never know what the other models can do. But I still feel attached to Opus, but this can be fammillarity. If I only had Qwen maybe it would be effectively identical at the end of the day. Hard to say.
Re: Qwen3.6-Plus: Towards real world agents
#173Not really interested in using models hosted on alibaba cloud. Like Qwen local for it’s privacy, but I trust the privacy of Google/OpenAI/Anthropic more than alibaba.
I build custom harnesses (like many of us) and I genuinely think Anthropic will eventually sue their customers if they detect they are selling competing harnesses (competes with their vertically integrated offerrings).
I feel Alibaba and DeepSeek see themselves more as infra. No urge to control the stack and litigate competition out of existence.
Re: Qwen3.6-Plus: Towards real world agents
#174Earlier quoted context omitted.
> not an open weight model like they’ve become known for. Right, they state that they'll release "smaller" variants openly at some point, with few details as to what that means. Will there be a ~300B variant as with Qwen 3.5? The blog post doesn't say.
I'm not interested in adopting an inferior closed source weight from a geopolitical rival. The open source weights argument was the one thing China had going and that I was seriously cheering them on for. They could have been our saviors and disrupted the US tech giants - and if it was open, I'd have welcomed it. Now they show their true colors. They want to train models on our engineering to replace us, while simult…
z.ai models are open weights. GLM-5.1 is very close to Opus with obvious exception of session length.
Only academic models will be true open source as companies can't legally afford to disclose learning inputs.
In regards to "They want to train models on our engineering to replace us". Some software engineers in China can run circles around some of the best teams in Silicon Valley. Days of U.S. hegemony are over. I recommend you make peace and make friends.
Re: Qwen3.6-Plus: Towards real world agents
#175[flagged]
In my experience, whenever an LLM says “wait, but actually” or some variant thereof is when you need to step in before it goes totally off the rails.
Re: Qwen3.6-Plus: Towards real world agents
#176I've gone through about 500M tokens on this model already. They've got some free inferencing options (such as on openrouter) ... $0 is hard to beat and it's creating not-crap.
How can it be free? What do you mean? EDIT: Ah, I see. Some kind of promotion. Pretty cool.
Also the Qwen cli, their vibe coding agent, allows a thousand free requests a day. In practice it's about ~300M tokens for me
It's very generous
I know I could certainly pay anthropic $1500 a day for my use and they'd be delighted... I'd rather pay $0 ... Just a personal preference
Re: Qwen3.6-Plus: Towards real world agents
#177Earlier quoted context omitted.
How can it be free? What do you mean? EDIT: Ah, I see. Some kind of promotion. Pretty cool.
Also the Qwen cli, their vibe coding agent, allows a thousand free requests a day. In practice it's about ~300M tokens for me It's very generous I know I could certainly pay anthropic $1500 a day for my use and they'd be delighted... I'd rather pay $0 ... Just a personal preference
Hahaha
Re: Qwen3.6-Plus: Towards real world agents
#178[dead]
Re: Qwen3.6-Plus: Towards real world agents
#179Earlier quoted context omitted.
> not an open weight model like they’ve become known for. Right, they state that they'll release "smaller" variants openly at some point, with few details as to what that means. Will there be a ~300B variant as with Qwen 3.5? The blog post doesn't say.
I'm not interested in adopting an inferior closed source weight from a geopolitical rival. The open source weights argument was the one thing China had going and that I was seriously cheering them on for. They could have been our saviors and disrupted the US tech giants - and if it was open, I'd have welcomed it. Now they show their true colors. They want to train models on our engineering to replace us, while simult…
looking at the other replies I'm not seeing what I consider the most important rebuttal to this argument: there is no real "adopting" in this hybrid open/close space right now, the lock-in is minimal and as much as different corporations are trying to create a lock-in effect by closing down their tools and interfaces, they are not really succeeding
I can constantly jump from one provider to another, and to my local servers which are already able to run very useful models at reasonable hardware cost, and I intend to continue doing that for the foreseeable future
the one thing I'm not going to do is tying my tooling to one provider or another or getting overtly used to the specifics of a model outside of my control
more than the weights or the training, which of course are very important, the real battle right now is for establishing some dependency mechanism so that your users won't just flee en-masse as soon as you inevitably try to abuse your market dominance and lock-in mechanisms, as is customary in everything computing these days - note that i don't explicitly talk about raking up prices, that is just one of the most difficult methods as people are very sensitive to that, when you can sneakily sell user data, get government contracts from never-disclosed conditions, or even just incorporate your intelligence to ad networks in one way or another
Re: Qwen3.6-Plus: Towards real world agents
#180Earlier quoted context omitted.
As a publicity stunt, releasing a 300B open model is pretty smart. You can talk about its strong performance and it being “open” and “available,” but it’s so large that most people can’t use it themselves and might try out the cloud-based offering.
The large models are actually MoE these days so they're usable on ordinary hardware with weights streaming from SSD, just very slow. You're nonethess right that it makes the cloud-based offering more popular, since you can use that for convenience after testing a few inferences locally.
“Usable but slow” is how you could run regardless of MoE or not, the model architecture has nothing to do with it. MoE might run N times faster than non MoE where N is the number of experts but that’s it.