Compare · ModelsLive · 2 picked · head to head
Claude Haiku 4.5 vs Grok 3
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Claude Haiku 4.5 wins on 7/8 benchmarks
Claude Haiku 4.5 wins 7 of 8 shared benchmarks. Leads in reasoning · math · coding.
Category leads
reasoning·Claude Haiku 4.5math·Claude Haiku 4.5knowledge·Grok 3coding·Claude Haiku 4.5
Hype vs Reality
Attention vs performance
Claude Haiku 4.5
#161 by perf·no signal
Grok 3
#150 by perf·#9 by attention
Best value
Claude Haiku 4.5
2.9x better value than Grok 3
Claude Haiku 4.5
12.4 pts/$
$3.00/M
Grok 3
4.3 pts/$
$9.00/M
Vendor risk
Who is behind the model
Anthropic
$380.0B·Tier 1
xAI
$250.0B·Tier 1
Head to head
8 benchmarks · 2 models
Claude Haiku 4.5Grok 3
ARC-AGI
Claude Haiku 4.5 leads by +42.2
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
Claude Haiku 4.5
47.7
Grok 3
5.5
ARC-AGI-2
Claude Haiku 4.5 leads by +3.9
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
Claude Haiku 4.5
4.0
Grok 3
0.1
FrontierMath-2025-02-28-Private
Claude Haiku 4.5 leads by +2.1
FrontierMath (Feb 2025) · original research-level math problems created by mathematicians, testing capabilities at the boundary of current AI mathematical reasoning.
Claude Haiku 4.5
5.9
Grok 3
3.8
FrontierMath-Tier-4-2025-07-01-Private
Claude Haiku 4.5 leads by +2.0
FrontierMath Tier 4 (Jul 2025) · the most challenging tier of frontier mathematics, containing problems that push the absolute limits of AI mathematical reasoning.
Claude Haiku 4.5
2.1
Grok 3
0.1
GPQA diamond
Grok 3 leads by +6.1
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Claude Haiku 4.5
61.6
Grok 3
67.7
MATH level 5
Claude Haiku 4.5 leads by +7.6
MATH Level 5 · the hardest tier of the MATH benchmark, featuring competition-level problems from AMC, AIME, and Olympiad-style mathematics.
Claude Haiku 4.5
96.4
Grok 3
88.8
OTIS Mock AIME 2024-2025
Claude Haiku 4.5 leads by +11.1
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Claude Haiku 4.5
66.6
Grok 3
55.5
WeirdML
Claude Haiku 4.5 leads by +8.2
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
Claude Haiku 4.5
45.4
Grok 3
37.2
Full benchmark table
| Benchmark | Claude Haiku 4.5 | Grok 3 |
|---|---|---|
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 47.7 | 5.5 |
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. | 4.0 | 0.1 |
FrontierMath-2025-02-28-Private FrontierMath (Feb 2025) · original research-level math problems created by mathematicians, testing capabilities at the boundary of current AI mathematical reasoning. | 5.9 | 3.8 |
FrontierMath-Tier-4-2025-07-01-Private FrontierMath Tier 4 (Jul 2025) · the most challenging tier of frontier mathematics, containing problems that push the absolute limits of AI mathematical reasoning. | 2.1 | 0.1 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 61.6 | 67.7 |
MATH level 5 MATH Level 5 · the hardest tier of the MATH benchmark, featuring competition-level problems from AMC, AIME, and Olympiad-style mathematics. | 96.4 | 88.8 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 66.6 | 55.5 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 45.4 | 37.2 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $1.00 | $5.00 | 200K tokens (~100 books) | $20.00 | |
| $3.00 | $15.00 | 131K tokens (~66 books) | $60.00 |
People also compared