TokenAtlasOpenAI vs Anthropic vs Gemini: API Pricing 2026
Flagship APIs priced head-to-head from the same maintained pricing catalog TokenAtlas uses in its calculator — kept in sync as vendors change rates.
Per-million token pricing
| Model | Provider | Input / 1M | Output / 1M | Context |
|---|---|---|---|---|
| GPT-4o | OpenAI | $2.500 | $10.000 | 128K |
| GPT-4o mini | OpenAI | $0.150 | $0.600 | 128K |
| GPT-4.1 | OpenAI | $2.000 | $8.000 | 1M |
| o1 | OpenAI | $15.000 | $60.000 | 200K |
| Claude 3.5 Sonnet | Anthropic | $3.000 | $15.000 | 200K |
| Claude 3.5 Haiku | Anthropic | $0.800 | $4.000 | 200K |
| Gemini 2.5 Pro | $1.250 | $10.000 | 1M | |
| Gemini 1.5 Pro | $1.250 | $5.000 | 2M |
Worked monthly cost: 10M input + 2M output
A modelled mid-volume SaaS workload — chat plus retrieval — gives concrete dollar figures, not just rate cards.
| Model | Modelled monthly cost | Context | Speed |
|---|---|---|---|
| GPT-4o | $45.00 | 128K | Fast |
| GPT-4o mini | $2.70 | 128K | Fast |
| GPT-4.1 | $36.00 | 1M | Fast |
| o1 | $270.00 | 200K | Slow |
| Claude 3.5 Sonnet | $60.00 | 200K | Medium |
| Claude 3.5 Haiku | $16.00 | 200K | Fast |
| Gemini 2.5 Pro | $32.50 | 1M | Medium |
| Gemini 1.5 Pro | $22.50 | 2M | Medium |
Which model should you pick?
- High-volume, cost-sensitive — GPT-4o mini (OpenAI) has the lowest combined list rate in this comparison.
- Long-context retrieval — Gemini 2.5 Pro (1M context) or Claude 3.5 Sonnet (200K context).
- Balanced general-purpose — GPT-4o at $2.5 input / $10 output per million tokens.
- Budget OpenAI tier — GPT-4o mini at $0.15 input / $0.6 output per million tokens.
Calculate your exact bill
Plug your own input and output volumes into the calculator to see the difference between providers on your traffic, not ours.
Open the AI cost calculatorFAQ
- What is the cheapest OpenAI model in this comparison?
- GPT-4o mini at $0.15 per million input tokens and $0.6 per million output tokens, with a 128K context window.
- Is Claude 3.5 Sonnet more expensive than GPT-4o?
- Yes. Claude 3.5 Sonnet lists at $3 input and $15 output per million tokens, versus $2.5 and $10 for GPT-4o. The gap on your own traffic depends on your input/output mix — model it in the calculator.
- Which model is cheapest in this comparison?
- On combined list rates, GPT-4o mini (OpenAI) is the lowest-cost option here at $0.15 input and $0.6 output per million tokens. Gemini 2.5 Pro offers a 1M context window when you need more room.
- How do I calculate my actual monthly API cost?
- Use the AI cost calculator — pick a model, enter monthly input and output token volumes, and TokenAtlas returns the modelled monthly cost plus the cheaper-model delta. Rates come from a maintained pricing catalog; TokenAtlas does not connect to your provider accounts.