opencode [FEATURE]: Terminal Bench 2.0 evaluation

Feature hasn't been suggested before.

[x] I have verified this feature I'm about to request hasn't been suggested before.

Describe the enhancement you want to request

Hi OpenCode team! Thank you for a great product. It would be great for OpenCode marketing purposes to evaluate your harness in Terminal Bench 2.0. Perhay you evaluated it before?

Dec 03 '25 10:12 rodion-m

This issue might be a duplicate of existing issues. Please check:

#2104: Submit to Terminal-Bench - A previous request to submit OpenCode to the Terminal Bench leaderboard for performance comparison.

Feel free to ignore if your specific case requires something different.

Dec 03 '25 11:12 github-actions[bot]

We are working on our own benchmarks, but I'd refer you to this: https://github.com/sst/opencode/issues/2104#issuecomment-3215423015

Dec 03 '25 16:12 rekram1-node

@rekram1-node thanks for the answer! Same results as CC - it's very good actually. But it will be interesting to see OpenCode's score in Terminal Bench 2.0 on open source LLMs. @thdxr what about Terminal Bench 2.0? Is it still "academic puzzle"? (I don't know and really curious). Btw, I'm a founder of CodeAlive context engine, we're also looking for good benchmarks, we don't explore a new terminal bench yet.

Dec 09 '25 21:12 rodion-m