Everything about Grok 4.20 (Reasoning)

Grok 4.20 (Reasoning) at a glance

xAI released Grok 4.20 (Reasoning) on March 9, 2026. It has a reasoning mode: it can "think" before answering, which helps with multi-step problems but uses more tokens, and therefore more money. It is a closed model: you access it through an API, from xAI or one of the 14 providers listed on models.dev. That makes it 7 months old as of October 22, 2026.

We file it under mid-range models (34 models in this category on the site).

What do independent benchmarks say?

According to results published by Epoch AI (checked on October 22, 2026), Grok 4.20 (Reasoning) has a capability index of 74.6 out of 100, computed on 4 benchmarks from our basket.

It is not among the top 30 of the weekly ranking.

• GPQA Diamond (PhD-level science questions): 89.3% — the best tracked model, Gemini 4 Argon, reaches 96.4%
• OTIS Mock AIME (competition mathematics problems): 92.2% — the best tracked model, GPT-5.5 Pro, reaches 100%
• FrontierMath (very hard research-level mathematics): 44.9% — the best tracked model, GPT-6 Astra, reaches 93.7%
• Terminal-Bench (agent tasks in a terminal): 57.3% — the best tracked model, GPT-5.5, reaches 84.7%

Technical specifications

Its context window reaches 2,000,000 tokens, roughly 3,000 pages of text at 500 words per page: enough to load a whole book or a mid-sized codebase. A single answer can run to 30,000 tokens.

It accepts as input: text and image.

It can call tools (functions, APIs), which makes it a possible foundation for an AI agent.

How much does Grok 4.20 (Reasoning) cost?

Grok 4.20 (Reasoning) costs $1.25 per million input tokens and $2.5 per million output tokens (median of 14 providers, checked on October 22, 2026).

To put it in perspective: reading 20,000 tokens (about thirty pages) and writing 2,000 costs roughly $0.03; a thousand of those tasks, $30.

That is 1.8 times cheaper than the median output price of mid-range models ($4.38).

To estimate your own budget, use the agent cost calculator.

Who is it for?

Given its category, Grok 4.20 (Reasoning) is mainly aimed at a balance between quality and cost, for everyday uses that do not justify a frontier model.

Its Terminal-Bench result (57.3%, above the median of tracked models) makes it a serious candidate for coding agents.

Its very large context window makes it worth considering for analysing large sets of documents in a single request.

Before adopting it, test it on your own examples: benchmarks measure standard tasks, not your job.

Alternatives

In the same category, the models closest to Grok 4.20 (Reasoning) are:

• Claude Opus 4.5 (Anthropic, index 74.3)
• GPT-5 (OpenAI, index 75.3)
• GPT-5.1 (OpenAI, index 75.8)
• Gemini 3.5 Flash (Google DeepMind, index 76)