This move today is really kind of interesting. $ZHIPU dumped directly in the middle of the session, down 9.7%. By all accounts, sentiment should have been pretty ugly, but after hours it flipped to the latest SpreadsheetBench V2 rankings—and the data actually looks very strong.
The overall score for Agent 2.0 in Decide is 33.81%, placing it directly above GPT-5.2, Gemini 3.1 Pro, GLM-5.0, DeepSeek V3.2, Kimi K2.5, Qwen3.5-397B-A17B, and MiniMax M2.5—basically overtaking leading models both at home and abroad in one sweep. This leaderboard is an independent global ranking focused specifically on spreadsheet performance, with very strong targeting; it’s not one of those vague, all-around benchmark tests. The “gold content” is still there.
Pay attention to one detail: GLM-5.0 is also on the list that’s been surpassed, and $ZHIPU itself is related to the Zhipu ecosystem. That means the parent company’s technological iterations are running ahead, yet the token price is still falling. This kind of divergence is quite common in retrospectives—news flow hasn’t caught up, capital gets dumped first, and then when the data is seen by more people, an expectation gap may emerge.
Of course, it’s still too early to talk about a reversal. A single-day drop of 9.7% suggests the selling pressure isn’t small, and the intraday rebound/containment is also not great; relying on just one leaderboard may not immediately change the pace. But from the fundamental anchor point, the performance ranking is a real competitive moat—at least it gives room for a later repair. Remember this score: 33.81%. Let’s see how the market digests it going forward. $ZHIPU
The overall score for Agent 2.0 in Decide is 33.81%, placing it directly above GPT-5.2, Gemini 3.1 Pro, GLM-5.0, DeepSeek V3.2, Kimi K2.5, Qwen3.5-397B-A17B, and MiniMax M2.5—basically overtaking leading models both at home and abroad in one sweep. This leaderboard is an independent global ranking focused specifically on spreadsheet performance, with very strong targeting; it’s not one of those vague, all-around benchmark tests. The “gold content” is still there.
Pay attention to one detail: GLM-5.0 is also on the list that’s been surpassed, and $ZHIPU itself is related to the Zhipu ecosystem. That means the parent company’s technological iterations are running ahead, yet the token price is still falling. This kind of divergence is quite common in retrospectives—news flow hasn’t caught up, capital gets dumped first, and then when the data is seen by more people, an expectation gap may emerge.
Of course, it’s still too early to talk about a reversal. A single-day drop of 9.7% suggests the selling pressure isn’t small, and the intraday rebound/containment is also not great; relying on just one leaderboard may not immediately change the pace. But from the fundamental anchor point, the performance ranking is a real competitive moat—at least it gives room for a later repair. Remember this score: 33.81%. Let’s see how the market digests it going forward. $ZHIPU