Models › Grok 4.7
SpaceXAI (formerly xAI) · Model details, evidence, and API facts
Grok 4.7 is SpaceXAI’s frontier model for agentic coding, long-horizon reasoning, and professional knowledge work
Starting provider API rate; context tiers apply
Grok 4.7 is SpaceXAI’s flagship model for agentic coding, long-horizon reasoning, and professional knowledge work, built on a larger base model
Grok 4.7 is a flagship AI model built by SpaceX AI
Only results mapped to this exact model and backed by a named source appear here. Different evaluation protocols are not treated as interchangeable
| Benchmark | Score | Evaluation details | Source |
|---|---|---|---|
Artificial Analysis Intelligence Index A composite of ten independent evaluations spanning agentic knowledge work, coding, science, and general reasoning. |
46 Composite index score |
Version: v4.3.2 · Includes AA-Briefcase, GDPval-AA, AutomationBench-AA, Terminal-Bench 4.0, SciCode, Humanity's Last Exam, and more Artificial Analysis Intelligence Index v4.3.2 incorporates 10 evaluations: AA-Briefcase v1.1, GDPval-AA v2.1, AutomationBench-AA, Terminal-Bench 4.0, SciCode, Humanity's Last Exam, GDP.pdf, CritPt, AA-Omniscience, AA-LCR v1.1 · Harness: Artificial Analysis standardized harness · Reasoning effort: xhigh | Methodology Artificial Analysis Independently reproduced |
AA Coding Agent Index End-to-end coding-agent performance combining DeepSWE, Terminal-Bench, and SWE-Atlas-QnA when run through Grok Build. |
56 Composite index score |
Version: Coding Agent Index · Harness: Grok Build (first-party coding agent) · Reasoning effort: xhigh · up from 47 with Grok 4.6 (xhigh), a nine-point gain broad-based across every component | Methodology Artificial Analysis Independently reproduced |
DeepSWE v1.1 A long-horizon software-engineering benchmark of real repository tasks graded by hand-written tests. |
73% Pass@1 |
Version: DeepSWE v1.1 · Harness: Grok Build · Reasoning effort: high · rises from 65% on Grok 4.6 to 73% on Grok 4.7 | Methodology Artificial Analysis Independently reproduced |
AA-Briefcase v1.1 A private benchmark for long-horizon agentic knowledge work modeled on real professional tasks. |
1657 Elo Elo rating |
Version: AA-Briefcase v1.1 · Harness: Artificial Analysis agentic harness · Reasoning effort: xhigh · gains +111 Elo over Grok 4.6 (high), placing it just behind Claude Opus 5 and Claude Fable 5.1 at the frontier | Methodology Artificial Analysis Independently reproduced |
GDPval-AA Evaluates models on economically valuable, real-world professional deliverables across industries. |
1695 Elo Elo rating |
Version: GDPval-AA v2.1 · Harness: Artificial Analysis agentic harness · Reasoning effort: xhigh · scores 1695 Elo on GDPval-AA, +90 ahead of Grok 4.6 (high) | Methodology Artificial Analysis Independently reproduced |
Arena scores come from blind human preference votes. They are reported separately from task benchmarks and are not used as substitutes for missing benchmark results
No exact Arena match is available for this model. Similar names and provider variants are not merged automatically.
Technical limits and reference provider prices come from the current catalog unless a reviewed external source is listed. Missing values stay marked as unavailable.
| Model | |
| Provider | SpaceXAI (formerly xAI) |
| Release date | September 21, 2026 |
| Catalog added | Sep 23, 2026 |
| Context window | 500,000 tokens |
| Maximum output | Not specified in reviewed sources |
| Knowledge cutoff | 2026-05 |
| License | Not available in reviewed sources |
| Capabilities and API | |
| Input types | File, Image, Text |
| Output types | Text |
| Reasoning | Supported (low, medium, high and xhigh; high is the default) |
| Provider endpoint tool parameters | Accepts tool parameters |
| Base provider API input | $2.00 / 1M tokens |
| Base provider cached input | $0.50 / 1M tokens |
| Base provider API output | $6.00 / 1M tokens |
| Provider tier at ≥ 200,000 prompt tokens | Input $4.00 / 1M · Output $12.00 / 1M · Cached input $1.00 / 1M |
Provider token rates are context-dependent. The base rates apply below the listed prompt thresholds; the matching tier applies at or above each threshold. These are reference provider API rates, not Pyxa AI checkout charges — access in Pyxa AI follows the allowances of your Pyxa AI plan.
Source links are attached to the facts and evaluations they support. Catalog-only values are not presented as independently verified claims
Open another model’s detail page or compare it directly with GPT-5.6 Sol. Suggestions come from the current catalog.
Use this model, then compare its response with other AI models in the same workspace.