Is Grok 4.7 available with Zero Data Retention on Elosia?

Grok 4.7 runs on Elosia's standard backend; ZDR is available on many other models in the catalog.

How much does Grok 4.7 cost on Elosia?

Grok 4.7 is billed by usage at $2 per 1M input tokens and $6 per 1M output tokens, within your Elosia plan.

What is Grok 4.7's context window?

Grok 4.7 supports a context window of up to 500,000 tokens.

Is Grok 4.7 hosted in the EU?

Grok 4.7 is not available on Elosia's EU-hosted (EdenAI) backend; it runs on the default backend.

When should you choose Grok 4.7?

Grok 4.7 is a flagship model, best for the most demanding reasoning, agentic and long-context tasks.

Grok 4.7

GrokFlagship

ThinkingTool UseVisionStructured Output

About this model

Grok 4.7 is SpaceXAI's flagship model for coding, agentic tasks, and knowledge work, succeeding Grok 4.6. It is particularly strong at long-running software engineering tasks, verifying its own work, and...

Performance Tier

Flagship

Grok 4.7 is a flagship model from Grok : the most capable in their lineup.

Best-in-class model from this provider. Highest performance across benchmarks, ideal for demanding tasks.

Pricing

This model is included in Elosia plans

Moderate

Moderate cost. A balanced choice for regular use.

Type	per 1M tokens
Input (prompt)	$2.00
Output (completion)	$6.00
Cache read	$0.500

Capabilities

Context Length500K

Max Output Tokens450K

TokenizerGrok

Inputtext, image, file

Outputtext

Release DateSeptember 21, 2026

Benchmarks

General Intelligence

MMLU

Not reported

GPQA Diamond

Not reported

Mathematics

MATH-500

Not reported

Programming

HumanEval

Not reported

SWE-bench Verified

Not reported

Reasoning

IFEval

Not reported

Humanity's Last Exam

43.1%

Where does Grok 4.7 stand?

Compare its performance index against every other model.

View the performance leaderboard

Recommended Use Cases

CodingAnalysisResearchGeneral ChatMathematics

Strengths

xAI flagship built for coding, agentic tasks and knowledge work, trained on a larger base model with longer reinforcement learning
Designed for long-running software engineering tasks, with explicit training to verify its own work
500K-token context with text and image input, function calling, web and X search, and code execution
Four selectable reasoning-effort levels, from low to xhigh, to trade depth against latency
Improved document and presentation creation for knowledge-work deliverables

Limitations

xAI publishes no standard academic benchmarks (no GPQA Diamond, SWE-bench or MMLU); the Humanity's Last Exam score shown comes from third-party testing
Proprietary, API-only model with undisclosed architecture; the faster variant is not available on the public API
Slow output generation at high reasoning effort, with long reasoning traces per task