← sources

Deepseek V4.1 flash inside Codex is amazing

source: reddit (reddit:LocalLLaMA)published 2026-09-23 13:56ingested 2026-09-24 02:40

retained text

Did anyone try Codex harness with Deepseek or other local models ? I like the reasoning show-off like GPT's ones, easier to steer and results are promising!

document fetched 2026-09-24 00:45 · canonical https://reddit.com/r/LocalLLaMA/comments/1wo6qcf/deepseek_v41_flash_inside_codex_is_amazing