Gemini 3 Pro vs Grok Code Fast 1

Benchmarks, API pricing and specs, head to head. Data updated 2026-06-10.

Gemini 3 Pro

Google · Nov 2025

73.6

Google's November 2025 frontier breakout — 91.9% GPQA Diamond and 37.5% HLE made it the reasoning leader of its generation.

Grok Code Fast 1

xAI · Aug 2025

A speed-first coding model built for agentic IDE loops, popular as a cheap fast tier in Cursor and similar tools.

The verdict

Gemini 3 Pro wins 1 of the 1 benchmarks these models share, against 0 for Grok Code Fast 1. On price the gap is dramatic: Grok Code Fast 1 works out roughly 8.6x cheaper per blended million tokens. Gemini 3 Pro also takes 1M of context versus 256K for Grok Code Fast 1.

Benchmark head-to-head 10

SWE-bench Verified
76.2%70.8%
Gemini 3 ProGrok Code Fast 1

Specs & pricing

Gemini 3 ProGrok Code Fast 1
modhub Index73.6
Input price / 1M$2$0.2
Output price / 1M$12$1.5
Context window1M256K
Max output66K
Open weightsnono
Reasoning modelyesyes
Multimodal inputtext, image, audio, videotext
Knowledge cutoffJan 2025Aug 2025
ReleasedNov 2025Aug 2025
Example monthly cost*$38.00$4.25

* 10M input + 1.5M output tokens per month at list prices, no caching. Green = better value on that row.

Frequently asked questions

Which is better, Gemini 3 Pro or Grok Code Fast 1?
Gemini 3 Pro wins 1 of the 1 benchmarks these models share, against 0 for Grok Code Fast 1. On price the gap is dramatic: Grok Code Fast 1 works out roughly 8.6x cheaper per blended million tokens. Gemini 3 Pro also takes 1M of context versus 256K for Grok Code Fast 1.
Which is cheaper, Gemini 3 Pro or Grok Code Fast 1?
Gemini 3 Pro costs $2/$12 per million input/output tokens, while Grok Code Fast 1 costs $0.2/$1.5. For a typical workload of 10M input and 1.5M output tokens per month, that's $38.00 versus $4.25.
Which model is better for coding, Gemini 3 Pro or Grok Code Fast 1?
On SWE-bench Verified — the standard agentic-coding benchmark — Gemini 3 Pro scores 76.2% versus 70.8% for Grok Code Fast 1, making Gemini 3 Pro the stronger pick for coding agents.

More comparisons