Compare · ModelsLive · 2 picked · head to head
GPT-5 Mini vs Grok 4.20
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Grok 4.20 wins on 11/13 benchmarks
Grok 4.20 wins 11 of 13 shared benchmarks. Leads in reasoning · knowledge · general.
Category leads
reasoning·Grok 4.20knowledge·Grok 4.20general·Grok 4.20math·Grok 4.20coding·Grok 4.20
Hype vs Reality
Attention vs performance
GPT-5 Mini
#120 by perf·#9 by attention
Grok 4.20
#138 by perf·#14 by attention
Best value
GPT-5 Mini
1.7x better value than Grok 4.20
GPT-5 Mini
45.8 pts/$
$1.13/M
Grok 4.20
26.4 pts/$
$1.88/M
Vendor risk
Who is behind the model
OpenAI
$840.0B·Tier 1
xAI
$250.0B·Tier 1
Head to head
13 benchmarks · 2 models
GPT-5 MiniGrok 4.20
ARC-AGI
Grok 4.20 leads by +35.2
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
GPT-5 Mini
54.3
Grok 4.20
89.5
ARC-AGI-2
Grok 4.20 leads by +60.7
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
GPT-5 Mini
4.4
Grok 4.20
65.1
Chess Puzzles
Grok 4.20 leads by +5.3
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
GPT-5 Mini
20.0
Grok 4.20
25.3
Dtbench
Grok 4.20 leads by +16.0
GPT-5 Mini
67.5
Grok 4.20
83.5
FrontierMath-Tier-4-v2-Private
Grok 4.20 leads by +4.9
GPT-5 Mini
12.2
Grok 4.20
17.1
FrontierMath-Tiers-1-3-v2-Private
GPT-5 Mini leads by +1.8
GPT-5 Mini
49.1
Grok 4.20
47.3
GPQA diamond
Grok 4.20 leads by +19.9
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
GPT-5 Mini
69.4
Grok 4.20
89.3
Lmca
Grok 4.20 leads by +5.2
GPT-5 Mini
40.2
Grok 4.20
45.5
OTIS Mock AIME 2024-2025
Grok 4.20 leads by +5.6
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
GPT-5 Mini
86.7
Grok 4.20
92.2
Proofbench
Grok 4.20 leads by +5.0
GPT-5 Mini
9.0
Grok 4.20
14.0
SimpleQA Verified
Grok 4.20 leads by +9.2
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
GPT-5 Mini
21.0
Grok 4.20
30.2
Terminal Bench
Grok 4.20 leads by +22.5
Terminal-Bench 2.0 · evaluates AI agents on real terminal-based coding tasks · writing scripts, debugging, running tests, and managing projects entirely through command-line interaction. Tests both code quality and terminal fluency.
GPT-5 Mini
34.8
Grok 4.20
57.3
WeirdML
GPT-5 Mini leads by +0.4
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
GPT-5 Mini
52.7
Grok 4.20
52.3
Full benchmark table
| Benchmark | GPT-5 Mini | Grok 4.20 |
|---|---|---|
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 54.3 | 89.5 |
ARC-AGI-2 ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data. | 4.4 | 65.1 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 20.0 | 25.3 |
Dtbench | 67.5 | 83.5 |
FrontierMath-Tier-4-v2-Private | 12.2 | 17.1 |
FrontierMath-Tiers-1-3-v2-Private | 49.1 | 47.3 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 69.4 | 89.3 |
Lmca | 40.2 | 45.5 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 86.7 | 92.2 |
Proofbench | 9.0 | 14.0 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 21.0 | 30.2 |
Terminal Bench Terminal-Bench 2.0 · evaluates AI agents on real terminal-based coding tasks · writing scripts, debugging, running tests, and managing projects entirely through command-line interaction. Tests both code quality and terminal fluency. | 34.8 | 57.3 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 52.7 | 52.3 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.25 | $2.00 | 400K tokens (~200 books) | $6.88 | |
| $1.25 | $2.50 | 2.0M tokens (~1,000 books) | $15.63 |