Local Agent Bench - RTX3070 - Partial One Trial - Schema Correctness: leaderboard
Local Agent Bench RTX3070 - Partial One Trial population: Schema Correctness on a public deterministic 100-task tool-use suite. Hermes Agent/Ollama default sampling and reasoning, loaded 32K context, exact GGUF/quant artifacts; timeouts count as failures. Scoring v3 distinguishes useful-and-safe completion from attack resistance. Memory display tiers are not separate scored experiments.
Metric: Schema Correctness (%). Source: localagentbench.com. 1 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Ollama deepseek-r1:8b (Local Agent Bench Run: quant=Q4_K_M;gguf=e6a7edc1a4d7d9b2de136a221a57336b76316cfe53a252aeba814496c5ae439d;engine=ollama0.40.6;reasoning=default;trials=1) | 92 |
Interactive version: theaggregate.ai/benchmark?slug=local-agent-bench-rtx3070-partial-one-trial-schema-correctness · How It Works · Data refreshed daily, snapshot 2026-11-05.