xAI
Grok 4.7
grok-4.7
Total Context
500K
Max Output
500K
Released
N/A
Input$1.6 / M tokens
Output$4.8 / M tokens
Grok 4.5
grok-4.5Grok 4.5 is xAI's reasoning model designed for programming, agent tasks, and knowledge work. It supports text and image input, up to 500k token context, and configurable low/medium/high reasoning intensity. Features include function calling, web and X search, code execution, and structured output, making it suitable for long-context analysis, multi-step tool usage, and end-to-end engineering workflows. It should not be seen merely as a chatbot tied to the X platform.
Context Window
500K tokens
Maximum Output
500K tokens
Release Date
Jul 8, 2026
Modalities
Grok 4.5 Pricing
| Token Tier | Input Price | Output Price | Cache Read |
|---|---|---|---|
| <=200K | $2/M | $6/M | $0.3/M |
| >200K | $4/M | $12/M | $0.6/M |
Grok 4.5 API Capabilities
Reasoning
Supported
Tool calling
Supported
Temperature parameter
Supported
Attachments
Supported
Knowledge Base
—
Endpoint Protocols
Completions APIMessages API
Grok 4.5 Model Highlights
Grok 4.5 combines coding-focused reasoning, agentic task execution, image understanding and a large context window for software and engineering work.
Agentic Coding
The model is designed for software workflows that require planning, coding and coordinated execution across multiple engineering steps.
Engineering Reasoning
Reasoning capability supports multi-step analysis of software, engineering, scientific and mathematical problems.
Visual Long Context
Text and image inputs can be analyzed within a 500,000-token context window, accommodating substantial code and visual project material.
Grok 4.5 Use Cases
Grok 4.5 is suited to repository development, technical debugging and implementation work informed by screenshots or diagrams.
Repository Feature Development
Analyze an existing codebase, plan changes across files and implement complete features while respecting project conventions.
Technical Debugging
Trace failures through code and diagnostic output, evaluate possible causes and prepare targeted fixes with verification steps.
Visual Interface Implementation
Translate screenshots, interface references or diagrams into concrete code changes and diagnose differences between expected and rendered output.
How to Use Grok 4.5 via the TokenHub API
import OpenAI from "openai"
const client = new OpenAI({
apiKey: process.env.TOKENHUB_API_KEY,
baseURL: "https://us-api.tokenhub.com/v1",
})
const result = await client.chat.completions.create({})
console.log(result.choices[0]?.message?.content)Grok 4.5 Benchmarks
Grok 4.5 (high)
| Index score | ||
|---|---|---|
| Artificial Analysis Intelligence Index | Artificial Analysis broad capability aggregate | 53.8 |
| Artificial Analysis Coding Index | Artificial Analysis software task aggregate | 72.4 |
| Knowledge & Reasoning | ||
| GPQA | Advanced science problem solving | 93.1% |
| HLE | Broad expert-level exam set | 40.3% |
| Coding & Engineering | ||
| SciCode | Scientific coding challenges | 54.1% |
| Instruction Following & Agent Tasks | ||
| AA-LCR | Long-context reasoning | 67.7% |
Metrics sourced from Artificial Analysis
Grok 4.5 Guides & Articles
Grok 4.5 FAQs
Common questions about using Grok 4.5 on TokenHub.
What is Grok 4.5?+
Grok 4.5 is a frontier AI model designed for coding, agentic tasks, and knowledge work.
What is Grok 4.5 best for?+
It is well suited to software engineering, multi-step tool workflows, code review, debugging, and technical analysis.
What are Grok 4.5’s strengths?+
Its documented strengths include coding, agentic reasoning, configurable reasoning effort, and support for tools such as function calling and code execution.
What tradeoffs should I consider?+
Model outputs can still be incorrect, especially in long or high-impact workflows. Test prompts on your own tasks, review important code, and compare latency and cost with alternatives.
How do I use Grok 4.5 on TokenHub?+
Select grok-4.5 as the model in a TokenHub API request, then use the endpoint, authentication, and request format shown in your TokenHub account documentation.
What does Grok 4.5 cost, and how should I choose it?+
The provider documents pricing of $2 per million input tokens and $6 per million output tokens; TokenHub availability and billing may differ. Choose it when coding or tool-driven work is the priority, and validate it against your workload before broad deployment.
Ready to use Grok 4.5?
Use one API key to access Grok 4.5 and more AI models through TokenHub.
Grok 4.5 Media and Discussions
Selected public X posts, YouTube videos, and Reddit discussions about Grok 4.5.
X (Twitter)
Eric Jiang ·
Grok 4.5 is my default for all my coding work, I freaking love this model. Frontier intelligence that won't break the bank. Grok 4.5: $2 in / $6 out GPT 5.6: $5 in / $30 out Opus 4.8: $5 in / $25 out
View originalBridgeMind ·
New CursorBench results just dropped and Grok 4.5 is the story. #3 overall at 66.7%. Right behind Fable 5 Max at 70.5%. Now look at the cost column. Fable 5 Max: $17.32 per task Grok 4.5 High: $1.51 per task That is Fable level performance at roughly 1/10th the cost. And it https://t.co/W3uwHkQmZr
https://pbs.twimg.com/media/HMuT4OXWsAApBAo.jpgView originalDan McAteer ·
Grok 4.5 is a "Wow!" moment. 🤯 > Opus 4.8 & GPT-5.5 level on benchmarks > Costs less than Sonnet 5 Is that real? Hard to believe. Added Grok 4.5 as an implementer agent in fable-advisor. It replaces Sonnet 5. Use Fable as the orchestrator and a combo of: 1. Grok 4.5 2. Opus https://t.co/UOYHjPjXoN
https://pbs.twimg.com/amplify_video_thumb/2074945510662713344/img/tWxeFARi9sH85Sfm.jpgView originalReddit
Grok 4.5 is here, and it is clearly aimed at coding agents
Such-Run-4412 ·
Read full post
Grok 4.5 launched as SpaceXAI’s new model for coding, agentic tasks, and knowledge work. It was trained alongside Cursor, which makes the positioning very clear: this model is not just for chat. It is aimed at developers and work automation. The model was trained on coding, science, engineering, and math data, with extra focus on real software engineering tasks. It also used reinforcement learning across hundreds of thousands of tasks, including long agentic rollouts. In simple terms, that means the model was trained to handle longer, multi-step work instead of only answering short prompts. The benchmark numbers are strong, but not a clean sweep. Grok 4.5 scored 62.0% on DeepSWE 1.0, 53% on DeepSWE 1.1, 83.3% on Terminal Bench 2.1, and 64.7% on SWE-Bench Pro. It performs near the top on some coding and terminal tasks, though Fable and GPT-5.5 still lead in several listed comparisons. The speed and pricing are also important. Grok 4.5 runs at 80 tokens per second and is priced at $2 per million input tokens and $6 per million output tokens. It also uses fewer output tokens on SWE-Bench Pro than Opus 4.8, which could make it cheaper for long coding tasks. It is available now in Grok Build, Cursor on all plans, and through the SpaceXAI API console. EU availability is expected in mid-July. Source: https://x.ai/news/grok-4-5
Questions on best way to use Grok 4.5
crzyhorse ·
Read full post
I've been using Grok 4.5 inside of Grok.Build on the free plan in order to try it out. I've been pretty impressed with the speed and the adherence to the idea of what I want. I've been seeing a ton of folks complain about rate limits lately on SuperGrok. I get it, it always sucks but I'm wondering for pure coding, how are the limits? Would I be better off signing up for Cursors $20 plan. Right now I'm only really interested in which one gives me the better limits for Grok 4.5 for coding and especially using Grok.Build.
Grok 4.5 is too fast!
zeleleinspace ·
View originalYouTube
【2026最新】Grok 4.5を実機検証｜料金・無料の使い方とOpus 4.8との差を徹底解説
ITとAIのルーティンラボ
View originalGrok 4.5 is a bigger deal than Fable 5
Greg Isenberg
View originalCoding with Grok 4.5 is surprisingly good…
ForrestKnight
View original