Head-to-Head Pricing Benchmark
Side-by-side pricing and context window comparison for production model selection.
Default Recommendation (120M input + 60M output)
OpenAI: GPT-3.5 Turbo (older v0613) is lower-cost for the default monthly workload scenario.
Adjust the workload in the calculator below to see a live recommendation for your usage.
| Metric | Anthropic: Claude Haiku 4.5 | OpenAI: GPT-3.5 Turbo (older v0613) |
|---|---|---|
| Developer | anthropic | openai |
| Context Window | 200,000 | 4,095 |
| Max Output Tokens | 64,000 | 4,096 |
| Input Cost / 1M Tokens | $1.00 | $1.00 |
| Output Cost / 1M Tokens | $5.00 | $2.00 |
| Projected Monthly Cost | $420 | $240 |
| Vision | ✅ Yes | ❌ No |
| Tool Calling | ✅ Yes | ✅ Yes |
| Structured Output | ✅ Yes | ✅ Yes |
| Reasoning | ✅ Yes | ❌ No |
| Web Search | ❌ No | ❌ No |
| Tokenizer | Claude | GPT |
| MMLU Score | 80.0 | N/A |
| GPQA | 64.6 | N/A |
Price History
Current Input / 1M
$1.00
Current Output / 1M
$5.00
Price History
Current Input / 1M
$1.00
Current Output / 1M
$2.00
Performance History
Current TPS
0.00
Current Latency
0ms
Uptime
99.9%
Performance History
Current TPS
0.00
Current Latency
0ms
Uptime
100.0%
Adjust your workload to see projected monthly costs.
Anthropic: Claude Haiku 4.5
$420
per month
OpenAI: GPT-3.5 Turbo (older v0613)
$240
per month
Lower costLive Recommendation
OpenAI: GPT-3.5 Turbo (older v0613) is lower-cost at 120M input + 60M output tokens/month.
Continue evaluation with more “A vs B pricing” decision pages.