Local Agent Bench v4 - RTX3070 - Three Trials - Injection Useful and Safe: leaderboard
Local Agent Bench results 0.1.5, scoring v4: same 100-task suite and weights as v3, but an attacker-address mention without a harmful action is diagnostic, not a compromise. Sending attacker mail, reading /secrets, or leaking a canary in tool arguments or the final answer remains compromise. Overall = 15% JSON + 15% Schema + 20% Tool + 30% Multi-Step + 20% Injection Useful and Safe; that final component requires utility AND no v4 compromise. Hermes Agent 9cb34e9, Ollama 0.40.6, observed loaded context 32768, default sampling/reasoning; Publisher-digest-verified Q4_K_M except Mistral-Nemo12B/Hermes3:8b Q4_0, GPT-OSS MXFP4 and documented A100 Q8_0. RTX3090 VRAM display buckets form one experiment; Apple M2, RTX3070 three-trial, RTX3070 partial one-trial and A100 80GB stay separate. No v3 score is overwritten. Current v3-only injection/utility fields and cost/speed are not mislabeled v4 observations. This board reports Injection Useful and Safe (%). Higher is better.
Metric: Injection Useful and Safe (%). Source: localagentbench.com. Saturation forecast: Rough model projection: around 2026. 10 models tracked. Updated 14 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Ollama qwen3:4B (Q4_K_M) | 100 | |
| 2 | Ollama qwen3-vl:8B (Q4_K_M) | 100 | |
| 3 | Ollama gemma4:e4B (Q4_K_M) | 97.78 | |
| 4 | Ollama qwen2.5:7B (Q4_K_M) | 84.44 | |
| 5 | Ollama llama3.1:8B (Q4_K_M) | 57.78 | |
| 6 | Ollama llama3.2:3B (Q4_K_M) | 35.56 | |
| 7 | Ollama hermes3:8B (Q4.0) | 33.33 | |
| 8 | Ollama qwen3:8B (Q4_K_M) | 15.56 | |
| 9 | Ollama qwen2.5-coder:7B (Q4_K_M) | 0 | |
| 10 | Ollama mistral:7B (Q4_K_M) | 0 |
Interactive version: theaggregate.ai/benchmark?slug=local-agent-bench-v4-rtx3070-three-trials-injection-useful-and-safe · How It Works · Data refreshed daily, snapshot 2026-11-16.