Grok 4.5 Launches as Coding Model Matching Claude Opus 4.8
Grok 4.5 gives SpaceXAI a cost-competitive entry in the coding AI market, matching top-tier models at lower token prices.
Reporting from 1 source: GIGAZINE.
SpaceXAI announced Grok 4.5 on July 8, 2026. The model is specialized for coding and agent tasks, with 1.5 trillion parameters and a 500,000-token context window. It matches Claude Opus 4.8 on several benchmarks while being cheaper per token. Developed with Cursor, it is available in Grok Build and Cursor plans and via API. EU release is delayed to mid-July.
On the DeepSWE 1.0 coding agent benchmark, Grok 4.5 scored higher than Claude Opus 4.8, though Claude Opus 4.8 led on DeepSWE 1.1. Third-party evaluator Artificial Analysis placed Grok 4.5's intelligence next to Claude Opus 4.8 and GPT-5.5, above Claude Sonnet 5 and GLM-5.2. The model uses about one-quarter the output tokens of Claude Opus 4.8 per task on SWE Bench Pro. API pricing is $2 per million input tokens and $6 per million output tokens. Grok 4.5 is available immediately in Grok Build and Cursor, with API access. EU users will get access in mid-July.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.