Interesting, I'm using it with Continue (with ollama) in code studio running locally. It's a bit slow but, tolerable and very good as far as the quality of code.
@@DigitalDesignET Very nice setup there! I agree that it's a great coding model. They just went a bit far in saying it's better than 3.5 Sonnet (old). The real contest for me is Deepseek 2.5 vs Qwen 2.5 Coder 32B! Video released soon!
@@ryun_d3v it makes more errors than the benchmarks indicate it should make. The quality difference vs 3.5 Sonnet is significant. For its speed, I'd continue with Deepseek 2.5 because they are similar in cost
Thank you! Very interesting.❤
Thank you @selub1058 for watching. Please subscribe if you haven't already
Interesting, I'm using it with Continue (with ollama) in code studio running locally. It's a bit slow but, tolerable and very good as far as the quality of code.
@@DigitalDesignET Very nice setup there! I agree that it's a great coding model. They just went a bit far in saying it's better than 3.5 Sonnet (old).
The real contest for me is Deepseek 2.5 vs Qwen 2.5 Coder 32B! Video released soon!
So what is your verdict?
@@ryun_d3v it makes more errors than the benchmarks indicate it should make. The quality difference vs 3.5 Sonnet is significant. For its speed, I'd continue with Deepseek 2.5 because they are similar in cost
@@ryun_d3v I'll make a Qwen 2.5 Coder 32B vs Deepseek 2.5 video very soon
Superb brother!
@@Tuguldurbayar thank you! Have a great day. Remember to subscribe please:)
so it worse than gpt 4o?
@@savasava9923 yep, from my uses of it. Instruction following is one the biggest difference