Cost by workload size
Price comparison
| Feature |  GPT-4o mini |  Gemini 2.5 Flash |
|---|---|---|
| Provider | OpenAI | |
| Input / M tokens | $0.15 | $0.30 |
| Output / M tokens | $0.60 | $2.50 |
| Cached input read | $0.07 | — |
| Context window | 128K | 1.0M |
| Batch discount | −50% | −50% |
| Vision | Yes | Yes |
| Audio | Yes | Yes |
| Reasoning tokens | — | Yes |
What your workload actually costs
Prices per token hide the real picture. Here are three common workloads computed with AITokenCalculator's own engine (medium reasoning effort, no batch, no cache):
| Workload | GPT-4o mini | Gemini 2.5 Flash | Cheaper |
|---|---|---|---|
| 1,000 in / 300 out | $0.00033 | $0.0018 | GPT-4o mini |
| 10,000 in / 2,000 out | $0.0027 | $0.013 | GPT-4o mini |
| 100,000 in / 20,000 out | $0.027 | $0.13 | GPT-4o mini |
| Monthly @ 1,000 req/day (standard) | $82.18 | $395.69 | GPT-4o mini |
Analysis
On input price, GPT-4o mini wins at $0.15/M — roughly 2.0× cheaper than its rival. Output tells a sharper story: GPT-4o mini charges $0.60/M , about 4.2× less than Gemini 2.5 Flash — and output is where chatty and agentic workloads bleed money.
Context capacity differs too: Gemini 2.5 Flash fits 1.0M tokens (~786,432 words) versus 128K. If you feed whole documents or codebases, that gap decides feasibility before price even matters.
The budget workhorses. Both are priced for scale; Gemini Flash usually offers the larger context window for document-heavy jobs, while GPT-4o mini matches OpenAI tooling and batch discounts. For pure cost-per-token at volume, check the scenarios below — margins are thin.
Numbers shift as providers reprice — check the full AI model pricing table, then run your own prompt through the calculator preloaded with GPT-4o mini or with Gemini 2.5 Flash.