SpaceXAI has released Grok 4.7, its new flagship model for coding, agentic tasks, and knowledge work. Grok 4.7 is built on a larger base model and a longer reinforcement learning run. It still ships at the same price and speed as Grok 4.6.

Is it deployable? Yes, as a hosted model. You can call grok-4.7 today through the xAI API, Cursor, Grok Build, OpenRouter, Vercel, and Cloudflare.

What Changed Under the Hood

SpaceXAI lists 4 changes over Grok 4.6:

1. A new, larger base model: Grok 4.7 does not reuse the Grok 4.6 base.
2. A longer RL run on harder tasks: The task mix is weighted toward problems that take many hours to complete.
3. Better self-verification and long-context handling: The company says the model checks its own work more carefully.
4. Native Grok Bot harness support: It was trained to understand the Grok Bot harness for conversational and knowledge work.

The developer docs list the API specs:

Benchmarks

The launch table compares Grok 4.7 at xHigh effort with Grok 4.6 High, GPT-5.6 Sol Max, and Fable 5.1 Max. The Grok 4.7 DeepSWE score was run at high effort. All scores are vendor-reported.

*High effort

Grok 4.7 improves on Grok 4.6 in every row. The largest jump is on Terminal-Bench 4.0, from 20.3% to 38.0%. EEBench rose 11 points to 64.0%, the top score in the table. On Harvey’s legal agent benchmark, Grok 4.7 scored 19.6% against 6.7% for Fable 5.1 Max.

Grok 4.7 does not lead across the board. Fable 5.1 Max tops 4 of 7 benchmarks, including a 57.9% Terminal-Bench score. GPT-5.6 Sol Max holds the top DeepSWE v1.1 result at 72.7%.

Price is the other axis. Fable 5.1 Max costs 5x more on input and about 8.3x more on output. GPT-5.6 Sol Max costs 2x more on input and about 3.3x more on output. On a CursorBench 4.0 cost-per-task chart, SpaceXAI places Grok 4.7 at the frontier in price-performance.

On GDPval, which tests professional knowledge work, Grok 4.7 xhigh scored 1,695 Elo. That is up from 1,605 for Grok 4.6 high. Fable 5.1 max leads at 1,735, and GPT-6 Astra max scored 1,542. SpaceXAI also says Grok 4.7 is better at creating documents and presentations.

Safety and Cybersecurity

Grok 4.7 ships with an entirely new safeguard stack. SpaceXAI calls it the strongest model it has tested on refusals and jailbreak resistance. It topped LatchBio’s biosafety benchmark at 62.4%.

On HackerBench v0.3, SpaceXAI’s own benchmark for risky and malicious cyber tasks, the model let 3.3% of risky dual-use prompts through. The company says it rarely blocks legitimate security work. Select cybersecurity partners now get invite-only access to its red-team capabilities for defense research.

Pricing and Availability

Grok 4.7 costs $2 per million input tokens and $6 per million output tokens. It is available in Cursor on all plans and is the default model in Grok Build. It is also served through the Grok API, OpenRouter, Vercel, and Cloudflare.

Grok 4.7 Fast is the same model on faster infrastructure, with twice the output speed at twice the price. The docs say it runs only in Cursor and Grok Build, not on the public xAI API. It is also excluded from Grok Build’s free tier.

A US regional endpoint at https://us.api.x.ai/v1 keeps inference in the United States at a 10% premium. SpaceXAI recommends setting a prompt_cache_key for reliable cache hits.

import os
from xai_sdk import Client
from xai_sdk.chat import user

client = Client(api_key=os.getenv("XAI_API_KEY"))
chat = client.chat.create(model="grok-4.7")
chat.append(user("Explain this repo."))
print(chat.sample().content)

Interactive Explainer

“;toks.forEach(function(t){h+=’

‘+t[0]+’

‘});
toks.forEach(function(tq,q){h+=’

‘+tq[0]+’

‘;
toks.forEach(function(tk,k){var a=ok(q,k);h+=”})});
$(“#grid”).innerHTML=h;
$$(“.rowl”).forEach(function(r){var f=function(){selq=+r.dataset.q;drawMask()};r.addEventListener(“click”,f);r.addEventListener(“keydown”,function(e){if(e.key===”Enter”||e.key===” “){e.preventDefault();f()}})});
var t=toks[selq],seen=toks.filter(function(_,k){return ok(selq,k)}).map(function(x){return x[0]}),seesX=seen.some(function(s){return s[0]===”X”});
var msg;
if(mode===”full”){msg=”

Everything sees everything

“+t[0]+” attends to all 9 tokens, including the noisy latent X. Because X changes at every step, the prefix keys and values change too, so nothing can be cached.

“}
else if(t[2]===”x”){msg=”

“+t[0]+” is the image being denoised

It attends to “+seen.join(“, “)+”. The latent reads the full prefix and every patch of its own block, so it gets complete context from text and references.

“}
else{msg=”

“+t[0]+” is part of the prefix

It attends to “+seen.join(“, “)+”. It “+(seesX?”sees”:”never sees”)+” the noisy latent X, so its keys and values stay identical across all denoising steps. That is what makes the prefix KV cache valid.

“}
$(“#mexp”).innerHTML=msg;size()}
$$(“#seg button”).forEach(function(b){b.addEventListener(“click”,function(){$$(“#seg button”).forEach(function(x){x.classList.remove(“on”)});b.classList.add(“on”);mode=b.dataset.m;drawMask()})});
drawMask();

/* ———- panel 4: resolution + rgba ———- */
var ars=[[“1:1”,2048,2048],[“4:3”,2400,1792],[“3:4”,1792,2400],[“3:2”,2528,1696],[“2:3”,1696,2528],[“16:9”,2752,1536],[“9:16”,1536,2752]];
$(“#ar”).innerHTML=ars.map(function(a,i){return ‘‘+a[0]+’‘}).join(“”);
function setAr(i){var a=ars[i],w=a[1],h=a[2],m=210,s=m/Math.max(w,h);var f=$(“#frame”);f.style.width=Math.round(w*s)+”px”;f.style.height=Math.round(h*s)+”px”;
$(“#px”).textContent=w+” x “+h;$(“#mp”).textContent=(w*h/1e6).toFixed(2)+” MP”;$(“#lat”).textContent=(w/16)+” x “+(h/16)}
$$(“#ar .chip”).forEach(function(c){c.addEventListener(“click”,function(){$$(“#ar .chip”).forEach(function(x){x.classList.remove(“on”)});c.classList.add(“on”);setAr(+c.dataset.i)})});
setAr(0);
var star=$(“#frame svg path”),dot=$(“#frame svg circle”);
$$(“#bg button”).forEach(function(b){b.addEventListener(“click”,function(){$$(“#bg button”).forEach(function(x){x.classList.remove(“on”)});b.classList.add(“on”);var f=$(“#frame”),v=b.dataset.b;
f.classList.toggle(“checker”,v===”checker”);f.style.background=(v===”checker”)?””:(v===”alpha”?”#000″:v);
if(v===”alpha”){star.setAttribute(“fill”,”#fff”);star.setAttribute(“stroke”,”#fff”);dot.setAttribute(“fill”,”#fff”)}else{star.setAttribute(“fill”,”url(#g)”);star.setAttribute(“stroke”,”#2B2A7A”);dot.setAttribute(“fill”,”#fff”)}})});
$(“#copy”).addEventListener(“click”,function(){var t=”This is an RGBA image with transparency. . The image has alpha channel and the background is transparent.”;
var done=function(){$(“#copy”).textContent=”Copied”;setTimeout(function(){$(“#copy”).textContent=”Copy”},1400)};
if(navigator.clipboard&&navigator.clipboard.writeText){navigator.clipboard.writeText(t).then(done,done)}else{done()}});

/* ———- panel 5: ship ———- */
var V=[
[“ok”,”Allowed”,”The license grants a royalty-free, worldwide right to use, copy, modify, and distribute the model for non-commercial purposes, which it defines as research or evaluation.”],
[“stop”,”Needs a separate commercial license”,”Commercial use is not covered. Qwen asks teams to request a license at model-business@notice.qwencloud.com before shipping.”],
[“warn”,”Allowed for non-commercial use, with conditions”,”If you use the model or its outputs to build and release another AI model, show \u201cBuilt with Qwen\u201d or \u201cImproved using Qwen\u201d in its docs. You cannot use \u201cQwen\u201d as the primary product name.”],
[“warn”,”Allowed, with notices”,”Give recipients a copy of the license, mark files you changed, and keep the Qwen attribution notice in a Notice file. The non-commercial limit still applies.”]];
function verdict(i){var v=V[i];$(“#verdict”).innerHTML=”;size()}
$$(“.use”).forEach(function(u){u.addEventListener(“click”,function(){$$(“.use”).forEach(function(x){x.classList.remove(“on”)});u.classList.add(“on”);verdict(+u.dataset.u)})});
verdict(0);
size();
})();