Grok Build 0.2.105 defaults to Grok 4.5

// 4h agoPRODUCT UPDATE

Grok Build 0.2.105 defaults to Grok 4.5

Grok Build version 0.2.105 updates its default model to Grok 4.5, featuring selectable high, medium, and low reasoning effort. Additional updates include allowing the /btw command to work inside the minimal interface (grok --minimal), introducing a new /summarize command for instant session summaries, and adding local network connection support.

// ANALYSIS

Defaulting to Grok 4.5 with adjustable reasoning levels represents a major step forward in tailoring developer workflows directly in the CLI.

–Tiered reasoning levels allow developers to customize execution speed and depth according to task complexity.
–Interactive shortcuts like `/summarize` help quickly document context and resolve session outcomes.
–Parity improvements for the minimal interface mode reduce friction for terminal power users.

// TAGS

grok-buildclidevtoolgrok-4.5updateartificial-intelligence

DISCOVERED

4h ago

2026-07-18

PUBLISHED

4h ago

2026-07-18

RELEVANCE

7/ 10

AUTHOR

mark_k

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

OPEN SOURCE1h ago

Vercel releases Python AI SDK public beta

Vercel has launched the public beta of its AI SDK for Python, porting its popular TypeScript-based toolkit for building AI applications and autonomous agent loops. The provider-agnostic SDK features zero-configuration setup, streaming, tool calling, and structured outputs using Pydantic models.

OPEN SOURCE1h ago

ProofAgent-Harness stress-tests AI agent reliability

ProofAgent-Harness is an open-source testing infrastructure that evaluates AI agent reliability and security through adversarial, multi-turn interactions. By employing a multi-juror consensus scoring system, the framework measures performance across critical dimensions like tool schema quality and injection hardening.

UPDATE2h ago

Google has rebranded NotebookLM to Gemini Notebook and added a secure cloud computer to enable native code execution for advanced data analysis.

Google has officially rebranded its AI research assistant NotebookLM to Gemini Notebook. Along with the new branding, Google introduced a secure cloud computer that allows the assistant to natively write and run code, enabling users to perform advanced data analysis directly on their uploaded sources.