Gemini 3 Pro Preview (Low) vs. Grok 4 Fast (Reasoning)

Comparison between Gemini 3 Pro Preview (Low) and Grok 4 Fast (Reasoning) across intelligence, price, speed, context window and more.

For details relating to our methodology, see our Methodology page.

	Google	SpaceXAI
Intelligence
Intelligence Index	22*	18*
Humanity's Last Exam	29%	19%
CritPt	0%	3%
AA-Omniscience	2	−30
AA-LCR v1.1	74%	74%
Cost
Price per 1M Tokens	$1.74	$0.23
Input Price per 1M Tokens	$2.00	$0.20
Output Price per 1M Tokens	$12.00	$0.50
Cache Hit Price per 1M Tokens	$0.20
Token Use
Performance
Technical specifications
Context Window	1000k tokens~1,500 A4 pages of size 12 Arial font	2000k tokens~3,000 A4 pages of size 12 Arial font
Release Date	November 2025	September 2025
Knowledge Cutoff	January 2025
Reasoning	Yes	Yes
Input modality	Supports: text, image, speech, and video	Supports: text and image
Output modality	Supports: text	Supports: text
Open Source (Weights)	No	No

* Estimated

Intelligence

Artificial Analysis Intelligence Index · Higher is better

Speed

Output tokens per second · Higher is better

No data available

Cost per Task

Weighted average cost (USD) per Intelligence Index task · Lower is better

No data available

Intelligence

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v4.3.2 incorporates 10 evaluations: AA-Briefcase v1.1, GDPval-AA v2.1, AutomationBench-AA, Terminal-Bench 4.0, SciCode, Humanity's Last Exam, GDP.pdf, CritPt, AA-Omniscience, AA-LCR v1.1

No publicly available API

Estimate (independent evaluation forthcoming)

Artificial Analysis Intelligence Index v4.3.2 includes: AA-Briefcase v1.1, GDPval-AA v2.1, AutomationBench-AA, Terminal-Bench 4.0, SciCode, Humanity's Last Exam, GDP.pdf, CritPt, AA-Omniscience, AA-LCR v1.1. See Intelligence Index methodology for further details, including a breakdown of each evaluation and how we run them.

Artificial Analysis Intelligence Index by Open Weights / Proprietary

No publicly available API

Estimate (independent evaluation forthcoming)

Indicates whether the model weights are available. Models are labelled as 'Commercial Use Restricted' if commercial use is limited by conditions, and as 'Non-commercial' if the license prohibits commercial use.

Intelligence Index vs. Cost per Intelligence Index Task

Artificial Analysis Intelligence Index · Weighted average cost (USD) per Artificial Analysis Intelligence Index task

Most attractive quadrant

Pareto line

AA-Briefcase v1.1

GDPval-AA v2.1

Cost per Task (USD, Log Scale)

Weighted average cost per Intelligence Index task. Each evaluation’s cost is calculated from input, cache hit, cache write, reasoning, and answer token prices, divided by task count, and weighted by its Intelligence Index weight.

Benchmarks

Intelligence Evaluations

Intelligence evaluations measured independently by Artificial Analysis · Higher is better

No publicly available API

AA-Briefcase v1.1Updated

Agentic knowledge work, (Elo-500)/2000

GDPval-AA v2.1Updated

Agentic real-world work tasks, (Elo-500)/2000

AutomationBench-AA

Agentic SaaS workflows

Terminal-Bench 4.0

Agentic coding & terminal use

SciCodeUnder review

Coding

Humanity's Last Exam

Reasoning & knowledge

GDP.pdf

Professional document reasoning, All-pass

CritPtUnder review

Physics reasoning

AA-Omniscience Accuracy

Knowledge

AA-Omniscience Non-Hallucination Rate

1 - hallucination rate

AA-LCR v1.1

Long context reasoning

Harvey LAB-AA v1.1Updated

Legal agentic work, Hallucination-Gated All-Pass Rate

EnterpriseOps-Gym-AA

Agentic business operations

Terminal-Bench-Science 0.1New

Agentic scientific research workflows in a terminal

FrontierCode 1.1New

Agentic software engineering in open-source repositories

AA-AnalystAgent

Quantitative analysis on spreadsheets & documents

ITBench-AA

Kubernetes incident root-cause analysis

MMMU-Pro

Visual reasoning

MLCR-AA

Medical long context reasoning

While model intelligence generally translates across use cases, specific evaluations may be more relevant for certain use cases.

AA-Omniscience

AA-Omniscience Index

AA-Omniscience Index (higher is better) measures knowledge reliability and hallucination. It rewards correct answers, penalizes hallucinations, and has no penalty for refusing to answer. Scores range from -100 to 100, where 0 means as many correct as incorrect answers, and negative scores mean more incorrect than correct.

No publicly available API

Intelligence Index Comparisons

Intelligence Index vs. Cost per Intelligence Index Task

Artificial Analysis Intelligence Index · Weighted average cost (USD) per Artificial Analysis Intelligence Index task

Most attractive quadrant

Pareto line

AA-Briefcase v1.1

GDPval-AA v2.1

Cost per Task (USD, Log Scale)

Cost

Cost per Intelligence Index Task

Weighted average cost (USD) per Artificial Analysis Intelligence Index task, segmented by token type. Lower is better

No publicly available API

Cost to Run Artificial Analysis Intelligence Index

Cost (USD) to run all evaluations in the Artificial Analysis Intelligence Index

No publicly available API

The cost to run the evaluations in the Artificial Analysis Intelligence Index, calculated using the model's input, cache hit, cache write, reasoning, and answer token prices and the number of tokens used across evaluations (excluding repeats).

Pricing: Cache Hit, Input, and Output

Price (USD per M Tokens)

No publicly available API

Price per token for cached prompts (previously processed), typically offering a significant discount compared to regular input price, represented as USD per million tokens. The values shown here are the cache hit price; cache write and cache storage are billed separately and vary by provider — see "Cache pricing by provider" for detail.

Context Window

Context window: tokens limit · Higher is better

No publicly available API

Larger context windows are relevant to RAG (Retrieval Augmented Generation) LLM workflows which typically involve reasoning and information retrieval of large amounts of data.

Maximum number of combined input & output tokens. Output tokens commonly have a significantly lower limit (varied by model).

Frequently Asked Questions

Gemini 3 Pro Preview (Low) is more intelligent. Gemini 3 Pro Preview (Low) scores 22 (estimated), compared with Grok 4 Fast (Reasoning) at 18 (estimated) on the Artificial Analysis Intelligence Index.

Grok 4 Fast (Reasoning) is cheaper. Grok 4 Fast (Reasoning) costs $0.23 per 1M tokens, compared with Gemini 3 Pro Preview (Low) at $1.74 per 1M tokens (7:2:1 cache hit/input/output ratio).

Grok 4 Fast (Reasoning) has a larger context window. Grok 4 Fast (Reasoning) supports 2.0M tokens, compared with Gemini 3 Pro Preview (Low) at 1.0M tokens.

Gemini 3 Pro Preview (Low) vs. Grok 4 Fast (Reasoning)

Intelligence

Speed

Cost per Task

Intelligence

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index by Open Weights / Proprietary

Artificial Analysis Intelligence Index

Open Weights

Intelligence Index vs. Cost per Intelligence Index Task

Cost per Intelligence Index Task

Artificial Analysis Intelligence Index

Benchmarks

Intelligence Evaluations

Intelligence Evaluation Relevance

Artificial Analysis Intelligence Index

AA-Omniscience

AA-Omniscience Index

AA-Omniscience Index

Intelligence Index Comparisons

Intelligence Index vs. Cost per Intelligence Index Task

Cost per Intelligence Index Task

Artificial Analysis Intelligence Index

Cost

Cost per Intelligence Index Task

Cost per Intelligence Index Task

Cost to Run Artificial Analysis Intelligence Index

Cost to Run Artificial Analysis Intelligence Index

Pricing: Cache Hit, Input, and Output

Cache Hit

Context Window

Context Window

Context Window for RAG

Context Window

Frequently Asked Questions

Which is more intelligent, Gemini 3 Pro Preview (Low) or Grok 4 Fast (Reasoning)?

Which is cheaper, Gemini 3 Pro Preview (Low) or Grok 4 Fast (Reasoning)?

Which has a larger context window, Gemini 3 Pro Preview (Low) or Grok 4 Fast (Reasoning)?