Someone supposedly got qwen working through Shakespeare + Maple, but I wonder if it was actually able to call tools. Should probably figure that out first, because it might not.

Reply to this note

Please Login to reply.

Discussion

Qwen3 is spotty with tools at the moment. gpt-oss-120b is better with it

Would it be viable to support GLM 4.6? It's the best open-source coding model we've tried for Shakespeare, but I don't like funneling user data to the Chinese government.

Might take some time but we will take a look