Kimi K3.1 Pro: Moonshot Discloses Cyber-Attack Nerfs as Internal Benchmarks Leak

Episode 1 · Moonshot admits nerfing Kimi K3.1 Pro's cyber capabilities before open release (2026-10-24, 20 posts)

In the Kimi K3.1 Pro technical report, Moonshot disclosed that as the model grew more capable, cybersecurity content naturally crept into general mid-training corpora and RL environments, steadily boosting vulnerability research ability through post-training. The team therefore proactively scrubbed the relevant data and intentionally suppressed the open release's offensive/defensive cyber capabilities, stating that downstream cyber ability would otherwise rise substantially.

Confirmed

• The reduction is disclosed in the technical report; multiple authors including @teortaxesTex, @scaling01, and @bookwormengr relayed the same fact
• Internal benchmark xdev-bench-hard (pass@3): the un-reduced checkpoint scored 90.5, dropping to 81.0 after data removal
• ExploitBench score of 74.7 (GPT-6 Astra: 100); ExploitGym score of 20.7 (GPT-6 Astra: 42+)
• @scaling01 notes that even with the data retained, internal cyber evaluations would still trail GPT-5.6-Sol

Why it matters

• A rare case of a vendor explicitly admitting capability-specific weakening in a technical report, with direct implications for the cybersecurity research value of open-weight models
• The quantitative data makes the "safety alignment vs capability ceiling" tradeoff assessable: nearly 10 points separate the un-reduced checkpoint from the released version
• Users relying on open models for vulnerability research or security testing should expect K3.1 Pro's cyber performance at its suppressed level

• Moonshot reveals it stripped cyber-offense training data from open-weight Kimi K3.1 Pro — teortaxesTex · 2026-10-24
• Kimi K3.1 Pro ships nerfed on cyber: internal bench drops from 90.5 to 81.0 after data removal — teortaxesTex · 2026-10-24
• Kimi K3.1 Pro trails GPT-5.6-Sol on cyber eval after Moonshot stripped cyber training data — scaling01 · 2026-10-24
• Moonshot lays out open-weight principles: never ship weights that extend the cyber frontier — teortaxesTex · 2026-10-24
• Moonshot says K3.1 Pro's offensive cyber capability is deliberately capped below frontier — dejavucoder · 2026-10-24
• Moonshot explains its capped-cyber-capability strategy for K3.1 Pro open weights — cephaloform · 2026-10-24
• Moonshot deliberately curbed Kimi K3.1's cyber capabilities: 74.7 on ExploitBench vs GPT-6 Astra's 100 — bookwormengr · 2026-10-24
• Moonshot deliberately limits Kimi K3.1 cyber capabilities: 74.7 vs 100 on Exploitbench — bookwormengr · 2026-10-24
• Moonshot removed cyber data from Kimi training to curb downstream attack capability — AashaySachdeva · 2026-10-24
• Moonshot published K3.1 cyber capability scores and stripped cyber data from its open-weight release — cedric_chee · 2026-10-24
• Moonshot removed cyber-related training data from open-weight Kimi K3.1 Pro — nrehiew_ · 2026-10-24
• Moonshot deliberately weakened Kimi K3.1 Pro's hacking ability before open release: 90.5% to 81% — rohanpaul_ai · 2026-10-24
• Moonshot strips cyber capabilities from open-weight Kimi, keeps full version internal — robleclerc · 2026-10-24
• Moonshot says it deliberately nerfed Kimi K3.1 Pro's cyber skills before open-sourcing — AccBalanced · 2026-10-24
• Kimi K3.1 open-weight release drops cyberattack training data — zxyzyxz · 2026-10-25
• Kimi K3.1's open-weights cyber pacing sparks debate on where to set the openness threshold — eliebakouch · 2026-10-25
• Not nerfing the cyber capabilities would have earned Kimi way more PR, researcher quips — eliebakouch · 2026-10-25
• K3.1 Pro becomes first open-weights release at frontier cyber capabilities, sparking debate — eliebakouch · 2026-10-25
• Moonshot strips cyberattack training data from latest Kimi model over hacking-tool fears — Polymarket · 2026-10-25
• Moonshot says it removed cyberattack training data from latest Kimi model — user1508123 · 2026-10-25

Episode 2 · Kimi K3.1 Pro Pass@3 on Internal Network Exploit Benchmark Jumps from 33% to 81% (2026-10-24, 2 posts)

Moonshot's Kimi K3.1 Pro reportedly scored 81% pass@3 on the internal xdev-bench-hard network exploitation benchmark, up from 33%. The result was shared and reshared by ricklamers with the full BenchLM page attached.

• Kimi K3.1 Pro Jumps to 81% pass@3 on Cyber Exploit Benchmark, Up From K3's 33% — ricklamers · 2026-10-24
• Duplicate Link Repost of Kimi K3.1 Pro xdev-bench-hard Benchmark Tweet — ricklamers · 2026-10-24