Sunday, November 15, 2026

Top 5 This Week

Related Posts

Grok 4.5: xAI breaks prices to compete with Claude


The essentials in 30 seconds

  • → xAI launched Grok 4.5 on July 8, 2026, which Elon Musk describes as “a model close to Opus, but faster and cheaper”.
  • → The shocking argument is the price: $2 in, $6 out per million tokens, much less than its rivals.
  • → It aims for efficiency for agents and repetitive tasksnot maximum raw performance.
  • → Gray area: unavailable in the European Union at launch, against the backdrop of compliance with theAI ActThe 2024 European regulation which regulates the development and use of AI according to risk levels..

In the race for AI, there are those who aim for the best model, and those who aim for the best quality-price ratio. With Grok 4.5, launched on July 8, 2026, xAI clearly chooses the second camp. Elon Musk does not claim to beat the best, but to get close to them for much less money. An assumed strategy, with its real strengths and its equally real limits.

July 8 · release
$2 / $6 · the price APIApplication Programming Interface, the programmatic access channel to a service or model.
Agents · the target
Outside the EU · at launch

01, The Promise

“Close to Opus, but cheaper”

Elon Musk did not oversell his model, and that is notable. He presents Grok 4.5 as “a model of the Opus class”, in reference to Claude Opus of Anthropic, before specifying: “roughly comparable to Opus 4.7, but much faster”. In other words, not the best on the market, but a very good model, fast and economical.

It’s a pricing strategy. On the independent rankings, Grok 4.5 is behind the leaders (Claude Fable 5, Opus 4.8, GPT-5.6) in raw performance. But it keeps up with them on agent tasks, while consuming much fewer tokens for a comparable result, which, on a large scale, changes the bill.

It is an Opus-class model, but faster, more tokenThe minimum unit of text (character, subword, or word) that an AI model processes. API prices are measured in tokens.-efficient and lower cost.

02, The price

The argument that hurts

This is where Grok 4.5 hits hard. In API, it costs $2 input and $6 output per million tokens. For comparison, Claude Opus 4.8 is $5/$25. For intensive use, particularly agents that run continuously, the cost difference becomes considerable.

On the general public side, Grok remains accessible via the X Premium+ subscriptions ($40/month) and the SuperGrok range (from $30 to $300/month depending on the intensity of use). The model is also integrated with coding tools and the X ecosystem, with real-time access to platform publications.

03, For whom

The right choice, for the right use

Grok 4.5 is not for everyone. For the most demanding tasks, where quality takes precedence over cost, Anthropic or OpenAI models maintain the advantage. But for a specific category of uses, it becomes very relevant:

High volume agents which chain together thousands of calls: the cost per task is paramount, and Grok 4.5 excels at it.

Repetitive tasks classification, extraction, mass summary, where a good model is enough and where the bill counts.

Uses linked to X thanks to native access to the platform’s publications in real time.

04, Reserves

Gray areas to keep in mind

The picture would not be complete without its shadows. First, Europe: Grok 4.5 was not available in the European Union at launch, a delay likely related to AI Act compliance requirements. French users will have to wait, or go through circuitous access.

Then, the question of safeguards. Grok has a history of more permissive moderation than its competitors, claimed by xAI as a choice of “less censorship”, but which has led to serious slip-ups in the past. Added to this is a question about neutrality: previous versions tended to draw on publications by Elon Musk himself on sensitive subjects. An efficient and inexpensive model, therefore, but whose limits must be known before deploying it.

To go further: The Grok model sheet · Claude Fable 5 is back · OpenAI launches ChatGPT Work + GPT-5.6 · Compare AI models

Frequently asked questions

What is Grok 4.5?
This is the new AI model from xAI (Elon Musk’s company), launched on July 8, 2026. Musk describes it as “close to Opus, but faster and cheaper”. It aims for efficiency and value for money, particularly for agents and repetitive tasks, rather than maximum raw performance.
How much does Grok 4.5 cost?
In API, it costs $2 in input and $6 in output per million tokens, much less than Claude Opus 4.8 ($5 / $25). On the general public side, it is accessible via X Premium+ ($40/month) and the SuperGrok range (from $30 to $300/month depending on usage).
Is Grok 4.5 better than Claude or GPT?
No on raw performance: on independent rankings, Grok 4.5 is behind Claude Fable 5, Opus 4.8 and GPT-5.6. Its advantage lies elsewhere: it follows the best in agent tasks while consuming fewer tokens, therefore for much less money.
Is Grok 4.5 available in France?
Not at launch. Grok 4.5 was not available in the European Union on July 8, a delay likely related to compliance with the AI ​​Act. French users will have to wait for a later European rollout.
For what uses should you choose Grok 4.5?
For high-volume agents (thousands of calls where cost per task takes precedence), repetitive tasks (classification, extraction, mass summary) and uses linked to X, thanks to its native access to the platform’s publications. For the most demanding tasks, Anthropic or OpenAI models remain ahead.
Does Grok pose moderation issues?
This is a point of vigilance. xAI claims more permissive moderation, which has caused Grok to seriously slip up in the past. Added to this is a question about neutrality, with previous versions tending to rely on Elon Musk’s publications on sensitive subjects. To know before any deployment.

Shredder Smith
Shredder Smith
Shredder Smith is the lead curator and digital persona behind topaitools4you.com, an AI directory dedicated to "shredding" through industry hype to identify high-utility software for everyday users. Smith positions himself as a blunt, no-nonsense reviewer who vets thousands of emerging applications to filter out overpriced "wrappers" in favor of tools that offer genuine ROI and practical productivity. The site serves as a watchdog for the AI gold rush, providing categorized rankings and transparent reviews designed to help small businesses and creators navigate the crowded tech landscape without wasting money on low-value tools.

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Popular Articles