aider Add more benchmarks

Add benchmarks for DeepSeek V3.1 and Qwen3-Coder

Aug 31 '25 11:08 markokocic

Can I take this task?

Sep 02 '25 00:09 AkshaySyal

I have not seen gpt5 on the leaderboard either?

Sep 02 '25 07:09 jakob1379

@jakob1379 👉 https://github.com/Aider-AI/aider/pull/4475

Sep 02 '25 15:09 butaca

GPT-5 is there now, but not gpt-5-mini

Sep 03 '25 20:09 markokocic

Would love to see the gpt oss models as well

Sep 09 '25 05:09 mattharrison

Other new DeepSeek models are missing now too.

DeepSeek-3.1-Terminus
DeepSeek-3.2 as well ase the latest Claude
Claude Sonnet 4.5

I know that model vendors report Aider Polyglot benchmark results themselves, but since those benchmars are missing from the official benchmarks, it's not clear if user can trust them.

Sep 30 '25 12:09 markokocic

GLM-4.6 and Claude Sonnet 4.5 benchmarks?

Other new DeepSeek models are missing now too.
* DeepSeek-3.1-Terminus

* DeepSeek-3.2
  as well ase the latest Claude

* Claude Sonnet 4.5
I know that model vendors report Aider Polyglot benchmark results themselves, but since those benchmars are missing from the official benchmarks, it's not clear if user can trust them.

deepseek benchies were merged

Oct 04 '25 09:10 Kreijstal