Grok Pricing
All plans and pricing as of 2026-09-28
X Premium
- ✓Higher query limits
- ✓Grok 4.20 access
- ✓Bundled X social features
SuperGrok
- ✓Full Grok 4.20 (4-agent multi-agent system)
- ✓DeepSearch mode
- ✓Highest rate limits
- ✓Think mode
- ✓$300/yr option (16% off)
SuperGrok Heavy
- ✓Grok 4 Heavy model
- ✓Highest priority
- ✓Multi-agent at scale
- ✓Note: Grok 4.3 beta-gating ended 2026-05-02
API (Grok 4.7, launched 2026-09-21)
- ✓500k context; cached input $0.50 per 1M
- ✓Long context (prompt at or above 200k tokens): $4 / $1 cached / $12 for every token in the request
- ✓Grok 4.7 Fast: 2x output speed at 2x price ($4 / $1 / $12; $6 / $1.50 / $18 above 200k) -- Cursor and Grok Build only, NOT on the public API, excluded from the Grok Build free tier
- ✓US regional endpoint (us.api.x.ai) bills 1.1x -- grok-4.7 and grok-4.6 only
- ✓Knowledge cutoff May 2026; Responses API always returns encrypted reasoning
- ✓Vendor benchmark charts only (CursorBench 4.0, DeepSWE v1.1, EEBench, AA Briefcase v1.1, Terminal-Bench 4.0, Harvey Legal, HealthBench Professional) -- no published table, numbers withheld here
API (Grok 4.5, launched 2026-07-08)
- ✓SpaceXAI + Cursor joint frontier MoE model
- ✓Faster variant $4/$18 via Cursor
- ✓Available in Grok Build, SpaceXAI console/API, Cursor (all plans)
- ✓NOT available in EU at launch
- ✓Vendor benchmarks: Terminal-Bench 2.1 83.3%, SWE-Bench Pro 64.7% (third-party verification pending)
API (Grok 4.3)
- ✓Production launch 2026-05-02 (~40% input / ~60% output price cut vs 4.20)
- ✓1M context window
- ✓Reasoning tokens billed at output rate
- ✓Native video input + PDF/PPT/spreadsheet output
- ✓Custom Voices voice cloning free on console (80+ presets, 28 languages)
- ✓Imagine Agent Mode (creative workflow agent, beta)
Is Grok Worth the Price?
Value Score: 7.5/10
Overall Score: 7.5/10 · People who live on X/Twitter and want an AI that can tap into that data in real-time. Also good for users who find mainstream chatbots too sanitized and want something with more personality.
Grok has come a long way from being dismissed as Elon's pet project. As of Grok 4.7 (2026-09-21) xAI ships a frontier-class coding and knowledge-work model at $2/$6 per 1M -- far below OpenAI and Anthropic list prices -- and places it in Cursor, Grok Build and GitHub Copilot on launch day, which makes the API the strongest part of the story. The consumer side is less clear-cut: at $30/mo for SuperGrok you are paying a premium for personality and real-time X data, and xAI still publishes its benchmark comparisons as charts rather than tables, so independent verification lags every launch. If real-time X data or cheap frontier tokens matter to you, Grok is a serious pick. If not, Claude or ChatGPT still give you more polish for the money.
The Tier List Tuesday
Weekly newsletter: tier movers, new entrants, and the VS of the week. Built from our daily AI-tool sweeps. No spam, unsubscribe anytime.
How Grok Pricing Compares
| Tool | Free Tier | Starting Price | Value Score | Overall |
|---|---|---|---|---|
| Grok(this tool) | Yes | $0 | 7.5/10 | 7.5 |
| Muse Spark (Meta) | Yes | $0 | 10/10 | 8.8 |
| Claude (Anthropic) | Yes | $0 | 8/10 | 8.5 |
| Gemini (Google) | Yes | $0 | 9/10 | 8.3 |
| MiMo (Xiaomi) | Yes | $0 | 9/10 | 8.3 |
| Hunyuan 3 (Tencent Hy3) | Yes | $0 | 9.5/10 | 8.1 |