Claude Haiku 4.5 vs GPT-5.4 Nano vs Gemini 3.1 Flash-Lite
Side-by-side pricing and capabilities from our model data. Prices verified against official pages.
Comparison
| Spec | Claude Haiku 4.5 Anthropic | GPT-5.4 Nano OpenAI | Gemini 3.1 Flash-Lite Google |
|---|---|---|---|
| Input ($/1M) | $1.00 | $0.200 | $0.250 |
| Output ($/1M) | $5.00 | $1.25 | $1.50 |
| Cached input ($/1M) | $0.100 | $0.020 | $0.025 |
| Context | 200K | 1M | 1M |
| Vision | Yes | No | Yes |
| Tier | budget | budget | budget |
| Batch API | Yes | Yes | Yes |
| Pricing | verified | verified | verified |
| Sample request (2K in / 600 out) | $0.005 | $0.001 ✓ | $0.001 |
At a typical 2K-in / 600-out request, GPT-5.4 Nano is cheapest at $0.001. GPT-5.4 Nano has the lowest input price ($0.200) and GPT-5.4 Nano the lowest output price ($1.25). Largest context window: GPT-5.4 Nano at 1M. Reads images: Claude Haiku 4.5, Gemini 3.1 Flash-Lite.
Want your own numbers? Use the cost calculator, the recommender, or see all comparisons.