Grok
Grok 4.3 — xAI
xAI's real-time, no-filter flagship. 1M token context, native video input, always-on reasoning with four intensity levels, and a price tag that undercuts everyone. Tops the leaderboards for agentic tool calling and instruction following.
Released
Apr 30, 2026
Context
1M tokens
Starting Price
$30/mo (SuperGrok)
API Pricing
$1.25 / $2.50
01
Overview
Grok is xAI's flagship AI assistant, built with a uniquely aggressive, unfiltered personality. Unlike other AI models that lead with corporate disclaimers, Grok gives you direct answers — sometimes with a sense of humor. Access it at grok.com or through the X (formerly Twitter) platform.
Grok 4.3 ships always-on reasoning with four configurable effort levels (none / low / medium / high), a 1M token context window, and a knowledge cutoff of December 2025. It's the first xAI model with native video input — processing video frames directly through a vision encoder — and tops public leaderboards for agentic tool calling and instruction following.
The 4.3 release also lands an aggressive ~40% cut on input pricing relative to the previous generation, making it one of the cheapest frontier-tier models per token. It's no longer beta — this is the model you ship on.
02
Key Capabilities
1M Token Context
Full 1,000,000-token context window with no output token limit
Native Video Input
First xAI model to process video frames directly through a vision encoder — no preprocessing required
Four Reasoning Levels
Configurable effort: none / low / medium / high (default low). Dial up for hard problems, dial down for speed
Real-Time X/Twitter Data
Native access to live public posts and trends
DeepSearch
Extended research mode scanning multiple web and X sources with citations
Top-Ranked Tool Calling
#1 on agentic tool calling and instruction following leaderboards. Built for code, agents, and multi-step orchestration
Custom Voices & Voice Library
Clone a voice from a short recording and manage it via the xAI console (paired with grok-voice-think-fast-1.0)
Aggressive Pricing
~40% cut on input tokens vs. prior generation. One of the cheapest frontier-tier models per token
03
Pricing
| Tier | Price | Details |
|---|---|---|
| Free | $0 | 10 prompts / 2 hours. No image generation. |
| SuperGrok | $30/mo | 100 default, 30 Think, 30 DeepSearch requests / 2 hrs. No X subscription needed. |
| X Premium+ | $40/mo | Highest usage caps + early access to new features. Includes X social features. |
| SuperGrok Heavy | $300/mo | Business/research tier. Largest quotas, earliest model access. |
| API (Grok 4.3) | $1.25 / $2.50 per 1M tokens | Input / output. Image input also $1.25/M. Generally available. |
| API (>200K tokens) | Higher tier rate | Requests over 200k total tokens billed at extended-context pricing. |
04
Best Use Cases
Real-time news and social media analysis
Financial forecasting and market analysis
STEM reasoning and academic research
Complex multi-perspective problem analysis
Tasks benefiting from live X/Twitter data
Getting unfiltered, direct answers without corporate hedging
05
Pro Prompts
Real-time Analysis
Give me the unvarnished latest on [current event]. Skip the caveats; I want signal, not noise.Financial Forecasting
Analyze the current market sentiment around [stock/sector] using real-time X data. Identify contrarian signals most analysts are missing.Research Deep Dive
Use DeepSearch to compile everything published in the last 48 hours about [topic]. Rank sources by credibility and highlight conflicting claims.Agentic Tool Calling
Set reasoning effort to high. Use whatever tools you have to check current pricing on [product] across [retailers], compare specs, and produce a buy/wait recommendation with citations.Video Analysis
Watch this clip and pull out: every product/feature shown, claims made on-screen, exact timestamps for each, and any contradictions between what's said and what's demonstrated.06
Limitations
High-effort reasoning mode meaningfully increases first-token latency — default low is fine for most tasks
Requests over 200K total tokens get billed at higher extended-context rates
Image generation lags Nano Banana 2 and DALL-E for typography-heavy work
Knowledge cutoff is December 2025 — rely on real-time X data and DeepSearch for anything newer
Image generation restricted to paid users only
8 legacy models retiring May 15, 2026 — existing API integrations need migration to 4.3
07
Multimodal Capabilities
Input
- Text
- Images (native vision encoder)
- Video (native, no preprocessing)
Output
- Text (no token output limit)
- Image generation (paid only, via Grok Imagine)
- Voice (grok-voice-think-fast-1.0 + Custom Voices)
Medical document analysis via photo upload
Can process circuit diagrams, design briefs, and visual inputs alongside text
08
Recent Updates
May 7, 2026
Custom Voices and Voice Library shipped — clone voices from short recordings via the xAI console
May 6, 2026
Grok 4.3 release announced via official email to all xAI API users
Apr 30, 2026
Grok 4.3 API rollout completed — 1M context, native video, four reasoning levels, 40% input price cut
Heads-Up
8 legacy xAI API models retiring May 15, 2026 — existing integrations should migrate to 4.3
Milestone
#1 ranking on CaseLaw v2 and CorpFin benchmarks, +300 Elo on GDPval-AA vs. Grok 4.20