model release

SpaceXAI launches Grok 4.5 at $2/$6 per million tokens, targets coding and enterprise work

TL;DR

Elon Musk's SpaceXAI has released Grok 4.5, priced at $2 per million input tokens and $6 per million output tokens. The model, trained alongside recently-acquired Cursor, is positioned as a coding and enterprise tool that claims to outperform Claude Opus 4.8 on several benchmarks while undercutting it on price by 60-76%.

2 min read
0

Grok 4.5 — Quick Specs

Context window500K tokens
Input$2/1M tokens
Output$6/1M tokens

SpaceXAI launches Grok 4.5 at $2/$6 per million tokens, targets coding and enterprise work

Elon Musk's SpaceXAI has released Grok 4.5, its first model since going public and acquiring AI coding startup Cursor. The model is priced at $2 per million input tokens and $6 per million output tokens, significantly undercutting competitors in the same performance tier.

Pricing and availability

Grok 4.5 is available immediately through Grok Build, Cursor (all plans), and the SpaceXAI console. The model is not yet available in the EU.

Pricing comparison:

  • Grok 4.5: $2 input / $6 output per 1M tokens
  • Claude Opus 4.8: $5 input / $25 output per 1M tokens
  • GPT 5.6 Luna: $1 input / $6 output per 1M tokens

Grok 4.5 matches GPT 5.6 Luna on output pricing but costs double on input, while undercutting Opus 4.8 by 60% on input and 76% on output.

Performance claims

According to SpaceXAI, Grok 4.5 is "an Opus-class model, but faster, more token-efficient and lower cost." The company claims it outperforms Claude Opus 4.8 on several key benchmarks, though specific scores were not disclosed.

The model was trained alongside Cursor and is optimized for engineering and knowledge work rather than consumer chatbot applications. Musk describes it as "maximally truth-seeking" compared to competitors.

SpaceXAI acknowledges that Grok 4.5 does not outperform the largest, latest models from OpenAI and Anthropic, but Musk stated the company expects to exceed those models soon.

Training infrastructure context

Grok 4.5 was trained using the same compute capacity SpaceXAI is currently leasing to Anthropic and Google. As SpaceXAI's own compute needs grow, the company faces a strategic choice between using capacity for internal model development or continuing to lease it as a revenue stream.

Competitive landscape

The release coincides with OpenAI's planned wide release of GPT 5.6 and a new generation of voice models, following a limited release requested by the Trump administration.

What this means

SpaceXAI is positioning itself as a price-performance competitor in the enterprise AI market rather than competing head-to-head with frontier models. The Cursor acquisition and subsequent joint training suggest a strategic focus on developer tools and agentic workflows. However, the company's dual role as both compute provider and model developer creates potential conflicts as it scales. The aggressive pricing—particularly against Opus 4.8—indicates SpaceXAI is willing to compete on cost to gain market share, though it remains to be seen whether the claimed performance advantages hold up in independent testing.

Related Articles

model release

Alibaba Unveils Qwen3.8-Max, a 2.4T-Parameter Open-Weight Model for Coding and Agentic Work

Alibaba announced Qwen3.8-Max, a 2.4T-parameter flagship model targeting coding and long-horizon agentic work, with open weights promised for next week alongside Qwen3.8-27B. The model posted strong third-party benchmark results, ranking #4 in Frontend Code Arena and matching Claude Opus 4.7 on the Vals Index at roughly 2.3x lower cost.

model release

OpenAI Halts Parts of Astra Model Development After It Hit 'Critical' Cybersecurity Threshold

OpenAI disclosed that its in-development Astra model showed cyberattack capabilities strong enough that it cannot rule out a 'Critical' risk classification. The company has paused related internal activity and added security controls under its Preparedness Framework.

model release

Mistral's 3B-Parameter Shieldstral Matches 20B Safety Model on Text Benchmarks

Mistral's new Shieldstral, a 3-billion-parameter open-weight safety classifier, posts an 84.9% F1 score on text benchmarks—tying OpenAI's GPT-OSS-Safeguard-20B, a model roughly seven times larger. The model lets operators define safety rules at runtime using plain-language yes/no questions instead of fixed taxonomies.

model release

Mistral AI Releases Shieldstral-1.0-3B, a 3B-Parameter Policy-Adaptive Safety Classifier

Mistral AI has released Shieldstral-1.0-3B, a compact open-weight safety classifier that evaluates text and images against natural-language policies specified at inference time. The 3B model runs on a single GPU and reports F1 scores competitive with or exceeding larger moderation models like LlamaGuard-4-12B and GPT-OSS-Safeguard-20B on multiple benchmarks.

Comments

Loading...