# Gemini 4 Neon walkthrough: why the same model swings from 39% to 12.8%

> Source: <https://dev.to/imapphelp/gemini-4-neon-walkthrough-why-the-same-model-swings-from-39-to-128-2k0e>
> Published: 2026-11-21 04:04:14+00:00

**Recommendation:** Run coding evals through the model's native tool-calling format rather than a generic function-calling shim, since rejected parallel tool calls can cut scores from 39% to 12.8%.
More in the playbook "Evals and testing for LLM apps": <https://wpnews.pro/playbooks/evals-testing>

I got tired of seeing the same model quoted at 39% and 12.8% with no explanation, so I recorded a full walkthrough breaking down exactly where those numbers come from. It started when I re-ran a coding eval and got a score nearly a third of what the leaderboard claimed — same weights, same prompts, different harness.

The video goes through the two evaluation setups side by side: one uses the model's native tool-calling format, the other wraps everything in a generic function-calling shim. That shim alone drops the score from 39% to 12.8% because the model keeps emitting parallel tool calls that the generic parser rejects, and every rejected call counts as a failed test case. I show the raw logs where you can see the parsing failures line up one-to-one with the score drop.

It cost me about six hours of GPU time re-running both configurations to confirm the delta wasn't noise, plus the editing time to make the logs readable on screen. The walkthrough is up at [https://www.youtube.com/watch?v=gBSZXtzBSfE](https://www.youtube.com/watch?v=gBSZXtzBSfE) if you want to see the exact log lines that explain the gap.
