Grok 4.5 Tops Terminal Agent Benchmark
XFreeze · x · 2026-07-15
Grok 4.5 secured first place on the Long-Horizon Terminal-Bench, outperforming Claude Fable 5, Claude Opus 4.8, and GPT-5.6-sol.
This benchmark tests whether an AI agent can continuously advance tasks without losing context during up to 90 minutes of terminal operations involving hundreds of dependent steps. The author notes that across 46 high-difficulty tasks and 18 frontier models, Grok 4.5 achieved the highest average reward of 0.505.
The post emphasizes that this demonstrates Grok 4.5's superior performance in complex agentic workflows, real-world coding automation, and difficult engineering tasks.
Related event: Grok 4.5 Tops Long-Horizon Terminal-Bench(3 posts)→
More from coding & agent
- OpenAI ships GPT-6.1 Terra; Brockman says its computer use now beats his own speed — gdb · 2026-10-27
- Kolab Benchmark Uses Pro Screen Recordings to Show AI Agents Overestimate Collaboration Quality — minnesotanlp · 2026-10-27
- Dev argues Claude burns far more tokens than GPT models for the same coding work — haider1 · 2026-10-27
- Stop building agents that can't change models: stable contracts, external state, real evals — rseroter · 2026-10-27
- Cheap models can cost more in agent harnesses than frontier ones, researcher warns — vasuman · 2026-10-27
- Kanzo: A 5k-Line Rust Kanban Binary With MCP for Claude Code and Codex — allmudi · 2026-10-27