Grok 4.5 is so cheap compared to Fable 5 and GPT 5.5 that benchmark gaps may not matter much
xAI launched Grok 4.5, a cost-efficient LLM trained on tens of thousands of Nvidia GB300 GPUs. It lags behind Fable 5 and GPT-5.5 in coding benchmarks but uses 4.2 times fewer tokens than Opus 4.8.
- Grok 4.5 is trained on tens of thousands of Nvidia GB300 GPUs, indicating significant computational resources.
- The model lags behind Fable 5 and GPT-5.5 in coding benchmarks but uses 4.2 times fewer tokens than Opus 4.8.
- Priced at $2 per million input tokens, Grok 4.5 is substantially cheaper than competitors like Fable 5 and GPT-5.5.
- EU availability is expected by mid-July, expanding its market reach in a regulated region.
xAI has released Grok 4.5, a new large language model trained on tens of thousands of Nvidia GB300 GPUs. While the model trails competitors like Fable 5 and GPT-5.5 in coding benchmarks, it compensates with significantly lower operational costs. Grok 4.5 requires 4.2 times fewer tokens than Opus 4.8, making it a more efficient option for users prioritizing cost over raw performance.
The pricing model is a standout feature, with input tokens priced at just $2 per million. This positions Grok 4.5 as a budget-friendly alternative in a market dominated by higher-cost models. Additionally, xAI has announced plans to expand availability to the EU by mid-July, broadening its user base in a key regulatory region.
The release highlights a growing trend in the AI industry where cost efficiency is becoming as critical as performance metrics. For developers and businesses, Grok 4.5 offers a compelling trade-off between affordability and functionality, particularly in applications where token usage is a major cost driver.
Source: Grok 4.5 is so cheap compared to Fable 5 and GPT 5.5 that benchmark gaps may not matter much. Read the full piece at the source.
Offers a cost-effective alternative for token-heavy applications, reducing operational expenses.
Provides a budget-friendly option for AI integration without sacrificing core functionality.
Demonstrates xAI's focus on cost efficiency, potentially influencing market dynamics and pricing strategies.
Showcases the growing importance of affordability in AI model adoption.
- GB300 GPUs
- Nvidia's high-performance GPUs designed for AI workloads, particularly large-scale training.
LLMSpaceXAI releases Grok 4.5, which Elon describes as an ‘Opus-class model’
GPT-Live, OpenAI's New Voice Model, Wants You to Talk Over It - Business Insider
OpenAI launches GPT-Live voice models that listen and speak simultaneously - Reuters
OpenAI to release its most powerful model after weekslong hold - Yahoo
OpenAI to publicly release GPT-5.6 AI models, ending government-requested limits - CNBC
AI ResearchI Built a Self-Improving AI, and So Can You
A Wired feature explores practical experiments where AI systems autonomously improve their own code and performance, challenging the dominance of top-tier labs.
Quoting Kenton Varda
A senior engineer halted AI-written pull request descriptions, citing they added no value and obscured code context.
AI ToolsIntroducing Claude apps gateway for AWS
Anthropic and AWS unveiled a self-hosted gateway to manage access, costs, and policies for Claude Code and Desktop on AWS infrastructure.
Biohub researchers use artificial intelligence to uncover new psoriasis targets - News-Medical
Researchers at Biohub have used AI to identify novel biological targets for psoriasis, potentially accelerating drug discovery.
RoboticsThis startup thinks robotics is about to have its ChatGPT moment
General Intuition raises millions to train robotics foundation models using video game data, aiming to reduce reliance on real-world robotics data.
AI ToolsGoogle AI Studio Adds ‘Import from GitHub’ to Build Mode, Turning an Existing Repo Into an Editable, Deployable App
Google AI Studio’s Build mode gains a GitHub import feature, letting developers turn existing repositories into editable, deployable applications with minimal setup.