Tech Meridian← ALL MODELS
PROMY MERIDIAN

XAI · MODEL RELEASE TRACKER

Grok 4.6

Grok 4.6 is xAI's flagship model (announced 2026-09-16) focused on long-running agents, interactive and visual work, and multi-step coding and knowledge tasks. It offers a 500K token context window and configurable reasoning effort levels. Grok 4.6 is available in Cursor and Grok Build and through major cloud marketplaces (Amazon Bedrock, Microsoft Foundry, Gemini Enterprise Agent Platform). It showed strong biosecurity refusal behavior on LatchBio’s benchmark and matches GPT-5.6 Sol on the Artificial Analysis Intelligence Index per xAI.

CURRENT SNAPSHOT4/5 DIMENSIONS WITH DATA

The dimensions that change the decision.

PRICING

Not established from the available sources.

CONTEXT WINDOW
Context window500,000 tokens (500K).
MODALITIES
Primary modalities / target domainsTargeted at coding, agentic knowledge work, and interactive/visual projects.
BENCHMARKS
Artificial Analysis Intelligence IndexxAI reports Grok 4.6 matches GPT-5.6 Sol on the Artificial Analysis Intelligence Index (a composite of multiple benchmarks).
AVAILABILITY
Product availabilityAvailable in Cursor and Grok Build; xAI offered 2× included usage inside Grok Build and Cursor for the first week of the release.
Cloud and platform availabilityAvailable via Amazon Bedrock, Microsoft Foundry, and Gemini Enterprise Agent Platform (Model Garden); available to developers in supported AWS Regions.

RELEASE TIMELINE

Published, source-backed release events only.

2 SOURCES · IMPORTANCE 8.0

SpaceXAI releases Grok 4.7 — faster, cheaper model for coding and knowledge work

SpaceXAI released Grok 4.7, a new base-model upgrade positioned as its most capable model for coding and knowledge work; it reportedly uses a larger base model, longer RL training focused on long-running tasks, and an entirely new safeguard stack. The company says Grok 4.7 matches the price and speed of Grok 4.6 while improving verification, longer-context handling, document/presentation generation, benchmark performance (CursorBench 4.0, GDPval, AA Briefcase), and safety metrics (LatchBio biosafety 62.4%, HackerBench v0.3 allowing 3.3% risky prompts); it is available in Cursor, Grok Build, the Grok API, third-party harnesses and cloud platforms, priced from $2 per million input tokens and $6 per million output tokens, with an optional faster (2x) variant at double the price.

→

VERIFIABLE FACTS

Every value stays attached to a source and date.

OTHER · Designed for long-running agents and complex multi-step tasksDEVELOPER CLAIM

Described as focused on long-running agents and ambitious interactive and visual work; stays with complex tasks across many steps (researching, analyzing, working across codebases, turning ideas into working artifacts).

MODALITIES · Primary modalities / target domainsDEVELOPER CLAIM

Targeted at coding, agentic knowledge work, and interactive/visual projects.

OTHER · Artificial Analysis Intelligence Index resultDEVELOPER CLAIM

Reported to match GPT-5.6 Sol on the Artificial Analysis Intelligence Index (a composite of nine benchmarks).

AVAILABILITY · Product availabilityDEVELOPER CLAIM

Available in Cursor and Grok Build; xAI offered 2× included usage inside Grok Build and Cursor for the first week of the release.

SAFETY · Biosecurity / refusal behaviorDEVELOPER CLAIM

On LatchBio's BioSecBench-Refusal suite, Grok 4.6 was reported as the strongest model tested at refusing disguised and hazardous biological tasks and was the only system to score above 50% on both measures.

WHAT CHANGED

Stored passport versions, without reconstructed history.

Passport updated13 facts
AVAILABILITY · Availability on Amazon Bedrock (packaging)− Launched on Amazon Bedrock (reported Bedrock launch date 2026-08-18); on Bedrock it is offered on both bedrock-mantle and bedrock-runtime endpoints and supports the Converse API alongside Chat Completions and Responses.
BENCHMARKS · Artificial Analysis Intelligence Index+ xAI reports Grok 4.6 matches GPT-5.6 Sol on the Artificial Analysis Intelligence Index (a composite of multiple benchmarks).
OTHER · Artificial Analysis Intelligence Index resultReported to match GPT-5.6 Sol on the Artificial Analysis Intelligence Index (a composite of nine benchmarks).→Reported to match GPT-5.6 Sol on the Artificial Analysis Intelligence Index (a composite of nine benchmarks).
SAFETY · Biosecurity / refusal behavior+ On LatchBio's BioSecBench-Refusal suite, Grok 4.6 was reported as the strongest model tested at refusing disguised and hazardous biological tasks and was the only system to score above 50% on both measures.
AVAILABILITY · Cloud and platform availability+ Available via Amazon Bedrock, Microsoft Foundry, and Gemini Enterprise Agent Platform (Model Garden); available to developers in supported AWS Regions.
OTHER · Configurable reasoning effort levelsSupports configurable reasoning effort levels: low, medium, high, and xhigh.→Supports configurable reasoning effort levels: low, medium, high, and xhigh.
CONTEXT WINDOW · Context window+ 500,000 tokens (500K).
CONTEXT WINDOW · Context window− Offers a 500,000-token (500k) context window.
AVAILABILITY · Availability in xAI products− Available in Cursor and Grok Build (announced as available today).
OTHER · Designed for long-running agents and complex multi-step tasksDescribed as focused on long-running agents and ambitious interactive and visual work; stays with complex tasks across many steps (researching, analyzing, working across codebases, turning ideas into working artifacts).→Described as focused on long-running agents and ambitious interactive and visual work; stays with complex tasks across many steps (researching, analyzing, working across codebases, turning ideas into working artifacts).
CAPABILITIES · Primary use cases+ Designed for long-running agents, ambitious interactive and visual work, and sustained multi-step coding and knowledge-work tasks; shows more self-testing and verification on longer trajectories.
AVAILABILITY · Product availability+ Available in Cursor and Grok Build; xAI offered 2× included usage inside Grok Build and Cursor for the first week of the release.
CAPABILITIES · Configurable reasoning effort levels+ Supports configurable reasoning efforts: low, medium, high, and xhigh.
RELEASE · Release announcementxAI announced the release of Grok 4.6 (announcement published 2026-09-16).→xAI announced the release of Grok 4.6 (announcement published 2026-09-16).
RELEASE · Release date+ Announced/released on 2026-09-16 (xAI announcement).
MODALITIES · Primary modalities / target domainsTargeted at coding, agentic knowledge work, and interactive/visual projects.→Targeted at coding, agentic knowledge work, and interactive/visual projects.
Passport created8 facts