Grok 4.5
READER SUMMARYA version-specific coding and agentic knowledge-work model released through the SpaceXAI API, Grok Build, and Cursor, with developer-reported gains across engineering benchmarks and long-running tool workflows.
Why this model scores 81.6
Primary launch evidence supports high coding, engineering, reasoning, office-tool, and agentic capability, including training on multi-hour rollouts. API access, low published pricing, Cursor distribution, and Grok Build make deployment broad. Strong cyber-adjacent coding and tool use create substantial misuse potential, while hosted access and the absence of an open-weight release keep deployment and residual control difficulty below freely downloadable frontier profiles.
News tied to Grok 4.5
The model score of 81.6 rates this model's risk profile. The overall Doom Index of 58.0 measures the accumulated evidence trajectory. These values answer different questions.
Each article's signed Doom Index movement is divided equally among the exact models named on that article. This prevents multi-model evidence from being counted in full on several model pages. The model's risk score itself never changes the overall index.
SpaceXAI releases Grok 4.5 for coding and agentic knowledge work
SpaceXAI released Grok 4.5 through its API, Grok Build, and Cursor, reporting stronger coding, engineering, and long-running agentic performance. The launch also priced broad developer access and positioned the exact version as the company's most capable model.
- Full item movement
- +3.19
- Grok 4.5 equal share
- +3.19
Model score history
- R1Doom Score 81.6
Initial model assessment from the dated primary Grok 4.5 launch source, filling one post-2020 historical gap.
11 Aug 2026