Grok 4 Fast
A proprietary cost-efficient Grok 4 variant with unified reasoning modes, native web and X search, code tools, and a two-million-token context window.
0 comments · 0 votes
Sign in to join the discussion →
No comments yet. Start the discussion.
Why this model scores 81.9
Developer benchmarks place capability near Grok 4, while end-to-end tool training and native multihop browsing create substantial autonomy. Free consumer access and inexpensive APIs justify maximal deployment reach. Dual-use cyber and CBRN ability plus tool access support elevated misuse, while refusal policies, input filters, prompt-injection mitigations, and agent evaluations lower residual control difficulty.
News tied to Grok 4 Fast
The model score of 81.9 rates this model's risk profile. The overall Doom Index of 66.8 measures the complete temporally weighted evidence record. These values answer different questions.
Each article's current Doom Index contribution is divided equally among the exact models named on that article. This prevents multi-model evidence from being claimed in full on several model pages. Model risk scores use a bounded temporal offset around their technical profile, but never feed back into the overall index.
Grok 4 Fast brings frontier search reasoning to free users
SpaceXAI released Grok 4 Fast to all Grok users and through low-cost APIs. The unified reasoning model combines a two-million-token context window with native web, X, and code-tool use, while its model card reports agent-abuse, prompt-injection, deception, and dual-use evaluations.
- Full item contribution
- +0.24
- Grok 4 Fast equal share
- +0.24
Model score history
-
R1
Doom Score 81.9
New exact public model verified from its dated launch page and model card.
14 Aug 2026