Claude Haiku 4.5 vs OpenAI GPT-5.4
Claude Haiku 4.5 is cheaper on output tokens, while OpenAI GPT-5.4 offers a larger context window. Choose Claude Haiku 4.5 or OpenAI GPT-5.4 based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Claude Haiku 4.5 | OpenAI GPT-5.4 |
|---|---|---|
| Provider | Anthropic | DigitalOcean |
| Input / 1M tokens | $1.00 | $2.50 |
| Output / 1M tokens | $5.00 | $15.00 |
| Context window | 200K | 400K |
| Parameters | — | — |
| Open weights | No | No |
| Released | Oct 2025 | Mar 2026 |
| LiveBench score | 61.0 | — |
| GPQA Diamond score | 71.2 | — |
| FrontierMath score | 5.9 | — |
| LMArena (Chatbot Arena) Elo score | 1397 | — |
| LMArena WebDev Elo score | 1330 | — |
| Epoch Capabilities Index score | 142 | — |
| SWE-bench Multilingual score | 64.7 | — |
| SWE-bench Verified score | 66.6 | — |
FAQ
Is Claude Haiku 4.5 or OpenAI GPT-5.4 cheaper?
Claude Haiku 4.5 is cheaper on output tokens ($5.00 vs $15.00 per 1M).
Which has the larger context window, Claude Haiku 4.5 or OpenAI GPT-5.4?
OpenAI GPT-5.4 has the larger context window (400K tokens).
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.