Grok 4.5 Tops Internal Benchmarks
billyuchenlin · x · 2026-07-11
Perplexity's Arav Srinivas expressed deep impression with Grok 4.5's performance. Key takeaways include:
- In the Computer harness, Grok 4.5 scored highest on the internal WANDR benchmark, which measures agentic research capabilities.
- It is priced at about half of Claude Opus 4.8.
- At a similar cost, it outperforms Perplexity's current GLM 5.2 post-trained model.
- Perplexity has deployed Grok 4.5 as the orchestrator model for Consumer Pro and Max users within Computer.
Related event: Grok 4.5 Joins Perplexity as Orchestrator, Tops WANDR Benchmark(10 posts)→
More from coding & agent
- Rowboat launches as an open-source, local-first AI coworker with memory — ycombinator · 2026-07-22
- Scoble says AI “loops” really means long-running multi-agent workspaces — Scobleizer · 2026-07-22
- Kimi Code opens a waitlist as Moonshot rolls out its coding product — Fabulous_Bonus_8981 · 2026-07-22
- Open-source runtime lets each repo define its own AI code reviewer — ibabufrik · 2026-07-22
- Indie Dev Asks: What's Actually Broken in Your AI Agent's Memory Today? — AcceptableTime7937 · 2026-07-22
- Fractal adds recursive agent loops for complex multi-step workflows — ryanpettry · 2026-07-22