| Usage Level | GPT-4o | Grok 2 | Savings |
|---|---|---|---|
| Light (100 calls/day) | $9.75 | $9.00 | Save $0.75 with Grok 2 |
| Medium (1K calls/day) | $180.00 | $168.00 | Save $12.00 with Grok 2 |
| Heavy (10K calls/day) | $2,250.00 | $2,100.00 | Save $150.00 with Grok 2 |
| Enterprise (100K/day) | $27,000.00 | $25,200.00 | Save $1,800.00 with Grok 2 |
Grok 2 is 4% cheaper overall. GPT-4o costs $2.5/M input vs Grok 2 at $2/M input.
At 800 input + 400 output tokens per call: GPT-4o costs $60.00 and Grok 2 costs $56.00.
With a proxy like Tokonomics, switching is a one-line config change. Route different features to different models and track costs per model automatically.