xAI · Grok 4

Grok 4.20

xAI's fourth-generation model with a 2M-token context window.

Grok 4.20, developed by xAI and released in 2023, is a large multimodal AI model designed to process and generate text, images, and file inputs. It features a massive 2,000,000-token context window, making it uniquely capable of handling extensive, complex inputs and tasks requiring intricate reasoning. Positioned as a flagship product, Grok 4.20 integrates advanced natural language processing, multimodal compatibility, and efficient large-scale data processing.

The model uses a transformer-based architecture optimized for cross-modality reasoning, allowing seamless synthesis between different input types. Trained on an up-to-date, diverse dataset and fine-tuned with reinforcement learning from human feedback, Grok 4.20 excels in maintaining coherence over long sequences and nuanced language handling, making it an ideal tool for developers addressing high-complexity needs.

Grok 4.20 represents the flagship tier of the Grok 4 series and introduces major advancements, including a doubling of the context window to 2,000,000 tokens and improved multimodal integration. These enhancements provide developers with more robust solutions for complex, large-scale data and multimodal problems.

Specs

Context window: 2M tokens
Input ($/1M tokens): $1.25
Output ($/1M tokens): $2.50
Modalities: Text · Image · File
Released: Mar 10, 2026
Weights: Closed

Pricing last synced May 9, 2026 via OpenRouter. Confirm against official docs before committing.

Capabilities

Tool use
Vision
Extended thinking
Prompt caching
Open weights

What it excels at

Massive context window
Processes up to 2,000,000 tokens, enabling seamless handling of lengthy documents or datasets.
Multimodal integration
Natively supports text, image, and file inputs, allowing cohesive cross-modality reasoning.
Transformer efficiency
Uses an optimized architecture for large-scale data processing and nuanced outputs.
Nuanced language handling
Excels at maintaining coherence and resolving ambiguity over complex and lengthy text inputs.
High input-output efficiency
Balances reduced latency with strong performance across demanding tasks.

When to use this model

→Document analysis and summarization - Handles large, complex documents with its extensive context processing.
→Customer support automation - Processes multimodal customer queries seamlessly for effective issue resolution.
→Long-form content generation - Generates detailed outputs across expansive inputs, like books or multi-document collections.
→Creative multimodal workflows - Merges text and images cohesively to support innovative projects.
→Multimodal data integration - Synthesizes diverse inputs into unified insights for enterprise-scale applications.

Analysis synthesized from gpt-4o, llama-4-maverick, phi-4, etc.

API model id

grok-4.20

Vendor docs: docs.x.ai

Compare Grok 4.20 with