New from xAI

Grok 4.5 online, pay per use

xAI released Grok 4.5 on July 8, 2026 as its flagship for coding, agentic tasks and knowledge work: a 500,000 token context window, text and image input, and the top score on agentic tool use in the independent Artificial Analysis index. On avots.ai you run it pay per use, with no X Premium or SuperGrok plan, and the same balance also covers Claude Opus 5, GPT-5.6, Gemini and 40+ image, video and music models.

Use Grok 4.5 now ✦ See pricing
Specs

Grok 4.5 at a glance

500K token context

Half a million tokens in one conversation, enough for a large codebase or a long document set.

2 USD in, 6 USD out

Per million tokens, with cached input at 0.50 USD, so repeated context gets cheaper.

93.1% GPQA Diamond

Graduate level science questions, with 83.3% on Terminal-Bench 2.1 and 64.7% on SWE-bench Pro.

Text and image input

Send screenshots, charts, diagrams and scanned pages, and get text back.

Effort you control

Reasoning effort low, medium or high, with high as the default, so you trade latency for depth.

Tools and JSON

Function calling and structured outputs work through the standard OpenAI compatible interface.

Head to head

Grok 4.5 next to Grok 4.3 and Claude Opus 5

 Grok 4.5Grok 4.3Claude Opus 5
Price per 1M in / out$2 / $6$1.25 / $2.50$5 / $25
Context window500K tokens1M tokens1M tokens
Terminal-Bench 2.183.3%not published89.1%
GPQA Diamond93.1%not published93.2%
Visionyesyesyes
On avots.aipay per usepay per usepay per use

Grok 4.5 sits between the cheap Grok 4.3 and the frontier Claude Opus 5: it matches Opus 5 on graduate science at a fraction of the output price, and it leads on agentic tool use. Switch between all three in the picker, one balance covers them.

What is new

What Grok 4.5 brings

Best in class tool use

It topped the agentic tool use category of the Artificial Analysis index at launch, with an overall score of about 54.

Terminal agents

83.3% on Terminal-Bench 2.1 and 64.7% on SWE-bench Pro, built for long shell and repository work.

Cheap cached context

Cached input at 0.50 USD per million tokens keeps repeated prompts and long agent loops affordable.

Reasoning you dial in

Set effort to low, medium or high per call and pay for depth only when the task needs it.

Try Grok 4.5 on avots
Three steps

How to use Grok 4.5

Create an account

Sign up on app.avots.ai with email, Google or Telegram. Takes a minute.

Top up from 0.99 EUR

1,000 tokens cost 0.99 EUR. No plan to choose, no renewal, the balance sits there until you use it.

Pick Grok 4.5

Choose it in the model picker, or let Agent Avots route for you. Every reply shows its token cost.

Sign up and start
One balance

Where you can use Grok 4.5

Web app

app.avots.ai on desktop and mobile, no install needed.

Telegram

Grok 4.5 in @AvotsAIbot, same balance, from any phone.

MCP

Use it inside Claude Desktop, Cursor or Cline via the MCP server.

Try Grok 4.5 for under a euro

1,000 tokens cost 0.99 EUR and a short Grok 4.5 answer costs a handful of them. The same balance unlocks Claude Opus 5, GPT-5.6, Gemini, Veo video and Nano Banana images, with no X Premium subscription.

Sign up and chat ✦
Guides

Related pages

Grok 4.6

Newer version. This model has a successor — same balance, same app.

Claude Opus 5

Anthropic's newest flagship, 1M context, 96% SWE-bench Verified.

GPT-5.6

OpenAI's newest family: Sol, Terra and Luna tiers, 1M context.

Kimi K3

Moonshot's frontier open model, number one at coding, 1M context.

Grok Imagine

xAI's image and video model for fast visual generations.

avots.ai runs 40+ AI models for chat, image, video and audio on one balance. See the full list on the Tools page or open the app and start creating.

Browse all model pages Open the app
FAQ

Grok 4.5 questions

What is Grok 4.5?
Grok 4.5 is xAI's flagship model for coding, agentic tasks and knowledge work, released on July 8, 2026. It has a 500,000 token context window, takes text and images as input, and supports function calling and structured outputs. It scores 93.1% on GPQA Diamond, 83.3% on Terminal-Bench 2.1 and 64.7% on SWE-bench Pro.
How much does Grok 4.5 cost on avots.ai?
Pay per use, no subscription. Grok 4.5 costs 2 USD per million input tokens and 6 USD per million output tokens, with cached input at 0.50 USD. Top up from 0.99 EUR for 1,000 avots tokens and every reply shows its cost before you run it.
What is Grok 4.5 best at?
Agentic tool use. On the independent Artificial Analysis Intelligence Index, Grok 4.5 scores about 54 and ranked fourth at launch, behind Claude Fable 5, Claude Opus 4.8 and GPT-5.5, but it topped the field on agentic tool use. Its 83.3% on Terminal-Bench 2.1 makes it strong for terminal agents and long tool chains.
Can I use Grok 4.5 without X Premium?
Yes. avots.ai routes to the real xAI model through official APIs, so there is no X Premium or SuperGrok plan to buy. You pay only for the tokens you use, starting at 0.99 EUR for 1,000 avots tokens.
Can I control how much Grok 4.5 thinks?
Yes. Grok 4.5 exposes a reasoning effort setting with low, medium and high, and high is the default. Lower effort returns answers faster and cheaper, higher effort spends more tokens on reasoning for hard problems.
Try it on avots ✦
Start creating