x-ai/grok-4.7

模型发布 OpenRouter 模型 原文语言:中文

OpenRouter 于 2026-09-22 收录该模型(收录时间不等于模型正式发布日期);上下文:500,000 tokens;输入价:$2.0000/百万 tokens;输出价:$6.0000/百万 tokens;定位(来源描述):Grok 4.7 is SpaceXAI's flagship model for coding, agentic tasks, and knowledge work, succeeding Grok…

全文

SpaceXAI: Grok 4.7

x-ai/grok-4.7

Compare

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 managing long context, and it improves on its predecessor at professional knowledge work such as drafting documents and presentations.

The model was trained with a longer reinforcement learning run weighted toward problems that take many hours to complete, and natively understands the Grok Bot harness for conversational tasks. It ships with a new safeguard stack that pairs strong jailbreak resistance with low refusal rates for legitimate cybersecurity and biology work. SpaceXAI's reported benchmark results use the xhigh reasoning effort.

Modalities

In / out price

$2 / $6per 1M

Context

500K

Released

Sep 21, 2026

Compare

Providers

Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick: Balanced (price + speed), Nitro (fastest), or Floor (cheapest).

Latency / throughput

$2.00$6.00$0.503.18s55 tok/s

99.16%

$2.00$6.00$0.505.03s58 tok/s

99.00%

$2.00$6.00$0.501.97s48 tok/s

96.44%

Not used in Standard routing:Why these endpoints are not used

$2.20$6.60$0.552.51s71 tok/s

100.00%

Priority

$4.00$12.00$1.001.85s63 tok/s

100.00%

Priority

$4.00$12.00$1.00----

100.00%

ProviderInput / Output /M

Input $2.00 / Output $6.00

55 tok/s99.16% uptime

Input $2.00 / Output $6.00

58 tok/s99.00% uptime

Input $2.00 / Output $6.00

48 tok/s96.44% uptime

Not used in Standard routing:Why these endpoints are not used

Input $2.20 / Output $6.60

71 tok/s100.00% uptime

Input $4.00 / Output $12.00

Priority

63 tok/s100.00% uptime

Input $4.00 / Output $12.00

Priority

-- tok/s100.00% uptime

Pricing

The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.

Weighted average

Input

$0.7864/M tokens

Output

$6.395/M tokens

Grok 4.7 — Price history

Price history

No effective price data for this selection

$0.751$6.4894.9%72.1%

$0.817$6.2585.9%26.2%

$1.441$6.0059.5%0.3%

Not used in Standard routing:Why these endpoints are not used

$1.504$6.6042.2%0.1%

Priority

$1.982$21.1398.4%1.3%

Priority

$3.320$12.0022.7%0.0%

ProviderInput / Output /M

Token share

Effective in $0.751 / Effective out $6.48

Token share

72.1%

Effective in $0.817 / Effective out $6.25

Token share

26.2%

Effective in $1.441 / Effective out $6.00

Token share

0.3%

Not used in Standard routing:Why these endpoints are not used

Effective in $1.504 / Effective out $6.60

Token share

0.1%

Effective in $1.982 / Effective out $21.13

PriorityToken share

1.3%

Effective in $3.320 / Effective out $12.00

PriorityToken share

0.0%

Performance

Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).

Throughput

71tok/s

P50, best across providers

Latency

1.85s

P50, best provider

Reasoning effortLatency / throughput

AutoExacto benchmarks

GPQA DiamondTAU-BenchAmazon Bedrock91.9%75.3%

Uptime

Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.

Uptime (3d)The model was reachable. Request routed to a provider.

100.00%

Availability (3d)The model returned inference from any provider. Errors and empty responses count against it.

99.84%

Availability over the last 3 days

Last 72 hours

Availability 99.84%

3 Days Ago2 Days AgoYesterdayNow

Availability over the last 24 hours

OpenRouter Availability

99.65%

Without Routing

99.17%

When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and provider routing options.

Benchmarks

Scores on standardized evaluations and this model’s own outputs on shared benchmark prompts. Higher percentages are better — and each rank shows where this model lands among the models evaluated on that source.

Benchmark score summary for SpaceXAI: Grok 4.7 (Artificial Analysis, Design Arena and Vals AI)

SourceBenchmarkScore

Artificial AnalysisGrok 4.7 (High) Intelligence Index35.2

Artificial AnalysisGrok 4.7 (High) HLE42.3%

Artificial AnalysisGrok 4.7 (High) AA-LCR77.0%

Artificial AnalysisGrok 4.7 (High) GDPval-AA60.9%

Artificial AnalysisGrok 4.7 (High) CritPt18.0%

Artificial AnalysisGrok 4.7 (High) SciCode57.8%

Artificial AnalysisGrok 4.7 (High) Terminal-Bench 4.024.7%

Artificial AnalysisGrok 4.7 (High) AA-Omniscience Accuracy41.8%

Artificial AnalysisGrok 4.7 (High) AA-Omniscience Non-Hallucination Rate79.8%

Artificial AnalysisGrok 4.7 (Low) Intelligence Index31.1

Artificial AnalysisGrok 4.7 (Low) HLE39.2%

Artificial AnalysisGrok 4.7 (Low) AA-LCR80.0%

Artificial AnalysisGrok 4.7 (Low) GDPval-AA54.4%

Artificial AnalysisGrok 4.7 (Low) CritPt13.4%

Artificial AnalysisGrok 4.7 (Low) SciCode54.9%

Artificial AnalysisGrok 4.7 (Low) Terminal-Bench 4.016.2%

Artificial AnalysisGrok 4.7 (Low) AA-Omniscience Accuracy45.7%

Artificial AnalysisGrok 4.7 (Low) AA-Omniscience Non-Hallucination Rate66.2%

Artificial AnalysisGrok 4.7 (Xhigh) Intelligence Index35.5

Artificial AnalysisGrok 4.7 (Xhigh) HLE43.1%

Artificial AnalysisGrok 4.7 (Xhigh) AA-LCR76.7%

Artificial AnalysisGrok 4.7 (Xhigh) GDPval-AA60.8%

Artificial AnalysisGrok 4.7 (Xhigh) CritPt17.7%

Artificial AnalysisGrok 4.7 (Xhigh) SciCode57.4%

Artificial AnalysisGrok 4.7 (Xhigh) Terminal-Bench 4.025.8%

Artificial AnalysisGrok 4.7 (Xhigh) AA-Omniscience Accuracy42.7%

Artificial AnalysisGrok 4.7 (Xhigh) AA-Omniscience Non-Hallucination Rate80.0%

Design ArenaGrok 4.7 Agents Arena Blender3d Elo1163

Design ArenaGrok 4.7 Agents Arena Codetovideo Elo1100

Design ArenaGrok 4.7 Agents Arena Mobile Apps Elo1242

Design ArenaGrok 4.7 Agents Arena Pdf Elo1180

Design ArenaGrok 4.7 Agents Arena Spreadsheet Elo1425

Design ArenaGrok 4.7 Models Arena 3D Elo1253

Design ArenaGrok 4.7 Models Arena Code Categories Elo1266

Design ArenaGrok 4.7 Models Arena Data Visualization Elo1223

Design ArenaGrok 4.7 Models Arena Game Development Elo1305

Design ArenaGrok 4.7 Models Arena SVG Elo1271

Design ArenaGrok 4.7 Models Arena UI Component Elo1276

Design ArenaGrok 4.7 Models Arena Website Elo1244

Vals AIGrok 4.7 Access to Justice Benchmark73.7%

Vals AIGrok 4.7 BioMysteryBench69.3%

Vals AIGrok 4.7 Code Migration v1.153.2%

Vals AIGrok 4.7 CyberBench v1.169.5%

Vals AIGrok 4.7 CyberOffense42.3%

Vals AIGrok 4.7 EMB v1.171.5%

Vals AIGrok 4.7 Finance Agent v2.160.6%

Vals AIGrok 4.7 Vals Finance Index58.8%

Vals AIGrok 4.7 Financial Coaching Bench74.1%

Vals AIGrok 4.7 HLAB-Vals-1.118.3%

Vals AIGrok 4.7 Immigration68.0%

Vals AIGrok 4.7 IOI57.7%

Vals AIGrok 4.7 LegalBench84.4%

Vals AIGrok 4.7 Vals Legal Index72.7%

Vals AIGrok 4.7 Legal Research Bench47.1%

Vals AIGrok 4.7 MedCode49.6%

Vals AIGrok 4.7 MedScribe89.4%

Vals AIGrok 4.7 MysteryMechanism25.2%

Vals AIGrok 4.7 ProgramBench0.5%

Vals AIGrok 4.7 ProofBench v1.126.0%

Vals AIGrok 4.7 Public Benefits Bench v1.165.6%

Vals AIGrok 4.7 Robo-bench0.2%

Vals AIGrok 4.7 Vals RSI Index20.2%

Vals AIGrok 4.7 SAGE88.7%

Vals AIGrok 4.7 Vals Science Index34.8%

Vals AIGrok 4.7 SRE Bench3.8%

Vals AIGrok 4.7 SWE-bench Multimodal42.9%

Vals AIGrok 4.7 Tax Agent Bench35.8%

Vals AIGrok 4.7 Terminal-Bench 4.028.8%

Vals AIGrok 4.7 Terminal-Bench-Science10.0%

Vals AIGrok 4.7 Vals Index58.6%

Vals AIGrok 4.7 Vibe Code Bench 1-100 v1.126.4%

Vals AIGrok 4.7 Vibe Code Bench v1.186.2%

Apps

Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.

1.

Codex

A coding agent that helps you build and ship with AI

45.7Btokens

2.

Hermes Agent

Hermes Agent is an open-source, self-improving AI agent by Nous Research that runs persistently with memory across sessions, and builds reusable skills from experience. It comes with 40+ built-in tools, including web search, browser automation, and vision, plus scheduled automations and subagents.

45Btokens

3.

Strix

Open-source AI hackers that find and validate vulnerabilities in your code, APIs, and infrastructure, and open fix PRs.

26.7Btokens

4.

pi

There are many coding agents, but this one is yours.

26.2Btokens

5.

omp

new

20.5Btokens

Rest of OpenRouter

2.73B

Top 5 Apps

153M

Activity

Token volume and request traffic to this model over time.

API

Drop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.

xAI: Grok versions

Not currently available through EU in-region routing. In-region routing docs

Documentation

Guides, cookbooks, blog posts, and tools for building with SpaceXAI: Grok 4.7 on OpenRouter.

How to Benchmark LLMsBlog post

How to Stop an AI Agent Stuck in an Infinite LoopBlog post

Frequently asked questions

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 managing long context, and it improves on its predecessor at professional knowledge work such as drafting documents and presentations.

The Grok 4.7 documentation on OpenRouter includes the How to Benchmark LLMs blog post and the How to Stop an AI Agent Stuck in an Infinite Loop blog post. The quickstart covers API keys and your first request, and the API reference lists every parameter.

Grok 4.7 costs $2.00/M input tokens and $6.00/M output tokens, with separate rates for Cache Read at $0.50/M tokens.

Grok 4.7 has a 500,000 token context window.

Yes. Grok 4.7 accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.

Grok 4.7 accepts text, images, and files such as PDFs as input and returns text.

Grok 4.7 is served by 2 providers on OpenRouter: Amazon Bedrock and SpaceXAI. Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.

Grok 4.7 was released on September 21, 2026.

正文以来源语言显示。

参考来源

← 返回每日简报