Grok Build July 13: Grok 4.5 Ships as the New Default at $2/$6 per 1M Tokens, CLI 0.2.96-0.2.98 Add Token Cost in /context and Queued Prompts

SpaceXAI Releases Grok 4.5: Musk's Opus-Class Model at $2/$6 per 1M Tokens, Default in Grok Build, Trained With Cursor

• On July 8-9, SpaceXAI released Grok 4.5, its strongest model yet.
• Musk called it an Opus-class model - faster, more token-efficient and lower cost, with roughly 2x token efficiency.
• Pricing is $2 per 1M input tokens and $6 for output - far below Opus's $5/$25 - with a 500K-token context window.
• It runs on the new 1.5-trillion-parameter V9 model, trained with Cursor.
• Grok 4.5 is the default in Grok Build and available in Cursor on all plans, but not yet in the EU (mid-July).
• The April SpaceXAI-Cursor deal may still resolve as a $10 billion investment or a $60 billion acquisition.

Editor’s read: For cost-sensitive agentic coding, Grok 4.5 is the headline value play this week: $2/$6 pricing and claimed 2x token efficiency undercut both Opus and GPT-5.6 Sol. Wait for EU availability if you are there.

Grok Build CLI 0.2.96 to 0.2.98: Token Cost in /context, Headless JSON Token Usage and Cost per Prompt, Voice Mode for API-Key Sessions, Queued Prompts

• Three Grok Build releases shipped this week.
• 0.2.96 (July 10) added live task activity in the tasks pane and per-repo MCP output truncation in .grok/config.toml.
• 0.2.97 (July 11) added headless JSON output with token usage and cost per prompt and session, Voice mode for API-key sessions, and auto-recovery for HTTP MCP servers after disconnects.
• 0.2.98 (July 12) pinned authentication to API keys or OIDC, surfaced token cost in /context, added queued prompts (type while the model works; press Enter twice to send immediately), and strengthened plan mode safeguards.

Editor’s read: The cost visibility is the useful thread across all three: token cost now shows in /context and in headless JSON per prompt and session - wire it into your CI to track spend on Grok 4.5's $2/$6 rate.

Grok 4.5 Now Available to All Vercel AI Gateway Customers, Announces Guillermo Rauch

• On July 8, Vercel CEO Guillermo Rauch announced that Grok 4.5 is now available to all Vercel AI Gateway customers.

Best Model for Coding After a Chaotic Week: Grok 4.5 vs GPT-5.6 vs Claude Fable 5 on Price and Benchmarks

• On July 10, Yellow.com compared the week's releases: Grok 4.5 lists at $2/$6 per million tokens, GPT-5.6 Sol at $5/$30, and Claude Fable 5 at $10/$50.
• Fable 5 leads coding at 80.4% on SWE-Bench Pro versus Grok 4.5's 64.7% and GPT-5.5's 58.6%, but independent analysis shows Grok 4.5 completing comparable coding tasks for $2.49 versus $11.80 for Fable 5.