GPT-4o mini vs Claude 3.5 Haiku: Benchmark Comparison
Detailed comparison of GPT-4o mini and Claude 3.5 Haiku covering benchmarks, pricing, context window, and compliance.
Key Specifications
| Specification | GPT-4o mini | Claude 3.5 Haiku |
|---|---|---|
| Vendor | openai | anthropic |
| Version | 4o-mini | 3.5-haiku |
| Release Date | 2024-07-18 | 2024-11-04 |
| Context Window | 128000 tokens | 200000 tokens |
| Input Modalities | text, image | text, image |
| Output Modalities | text | text |
| License | Proprietary | Proprietary |
| SOC2 | ✓ | ✓ |
| HIPAA | ✓ | ✗ |
| GDPR | ✓ | ✓ |
| ISO 27001 | ✓ | ✗ |
Benchmark Results
| Benchmark | GPT-4o mini | Claude 3.5 Haiku | Winner |
|---|---|---|---|
| ARC | 93.9 | 91.8 | GPT-4o mini |
| BBH | 70.3 | 70.2 | GPT-4o mini |
| GPQA | 42.6 | 31.1 | GPT-4o mini |
| GSM8K | 75.1 | 75.5 | Claude 3.5 Haiku |
| HUMANEVAL | 78 | 73.8 | GPT-4o mini |
| IFEVAL | 67.9 | 73.6 | Claude 3.5 Haiku |
| MATH | 51.8 | 53 | Claude 3.5 Haiku |
| MMLU | 79.7 | 77.6 | GPT-4o mini |
| MUSR | 53.1 | 52.3 | GPT-4o mini |
| WINOGRANDE | 79.6 | 83.8 | Claude 3.5 Haiku |
Pricing Comparison
| Tier (per Mtok) | GPT-4o mini | Claude 3.5 Haiku |
|---|---|---|
| Input | $0.15 | $0.8 |
| Output | $0.6 | $4 |
| Cache Read | $0 | $0 |
| Cache Write | $0 | $0 |
GPT-4o mini vs Claude 3.5 Haiku
Model Overview
GPT-4o mini and Claude 3.5 Haiku are both notable options in the AI model market. This page compares their benchmarks, pricing, and compliance.
Key Specifications
| Vendor | Release Date | Context Window | License |
|---|---|---|---|
| Openai / Anthropic | 2024-07-18 / 2024-11-04 | 128K / 200K | Proprietary / Proprietary |
Benchmark Performance
| Benchmark | GPT-4o mini | Claude 3.5 Haiku | Winner |
|---|---|---|---|
| ARC | 93.9 | 91.8 | A |
| BBH (BIG-Bench Hard) | 70.3 | 70.2 | Tie |
| GPQA | 42.6 | 31.1 | A |
| GSM8K (Grade School Math 8K) | 75.1 | 75.5 | Tie |
| HumanEval | 78.0 | 73.8 | A |
| IFEval | 67.9 | 73.6 | B |
| MATH | 51.8 | 53.0 | B |
| MMLU (Massive Multitask Language Understanding) | 79.7 | 77.6 | A |
| MUSR | 53.1 | 52.3 | A |
| WinoGrande | 79.6 | 83.8 | B |
Pricing Comparison
| Input | Output | Cache Read | Cache Write |
|---|---|---|---|
| — / — | — / — | — / — | — / — |
per million tokens — A / B
Strengths & Weaknesses
GPT-4o mini
- ✅ Input Modalities: text, image, audio.
- ⚠️ Proprietary, not self-hostable.
Claude 3.5 Haiku
- ✅ Reliable general-purpose model.
- ⚠️ Proprietary, not self-hostable.
Editor’s Take
GPT-4o mini and Claude 3.5 Haiku each have their strengths. Choose based on workload (code, long context, vision), referencing the tables above.
FAQ
Which model is better for coding tasks?
Refer to the HumanEval benchmark table; the model with a higher score is better suited for coding tasks.
Which model is cheaper?
Refer to the pricing comparison table above; the model with lower input/output prices is more cost-effective.
Which has a longer context window?
Refer to the key specifications table; the model with a larger context window is better for long documents.
References
Editor's Take
See Editor's Take section.