Gemini 3.1 Pro vs GPT-5: pricing & cost comparison
On input tokens, GPT-5 is the cheaper of the two — 38% less per million ($2 vs $1.25). On output, GPT-5 is 17% cheaper ($12 vs $10) — and since output is usually the dominant cost driver, that gap matters more than it looks.
Side by side
| Gemini 3.1 Pro | GPT-5 | |
|---|---|---|
| Input / 1M tokens | $2 | $1.25 |
| Output / 1M tokens | $12 | $10 |
| Context window | 1,000,000 | 400,000 |
| Token-count accuracy | ±3% | exact |
| Cost — 10,000 input + 2,000 output tokens | $0.044 | $0.0325 |
What a real request costs
Take a representative turn — 10,000 input + 2,000 output tokens. Gemini 3.1 Pro comes to $0.044, GPT-5 to $0.0325. Across 100,000 requests that's a $1150 swing in favour of GPT-5. To run the numbers on your actual prompt, paste it into the calculator and toggle Compare across all models.
Close call — decide on output cost
Gemini 3.1 Pro and GPT-5 are close enough on price that the tie-breakers matter more than the headline rate. Output tokens are usually the dominant cost driver, so weigh $12 vs $10 per 1M output first, then context window — Gemini 3.1 Pro carries the larger one at 1,000,000 tokens. They're different vendors, so factor in a small tokenizer calibration buffer on whichever side isn't exact.
See the full breakdown on the dedicated pages for Gemini 3.1 Pro and GPT-5.
FAQ
- Is Gemini 3.1 Pro or GPT-5 cheaper?
- For a typical request (10,000 input + 2,000 output tokens), GPT-5 is cheaper — about 26% less, or roughly $1150 saved per 100,000 requests. Gemini 3.1 Pro runs $2/$12 per 1M input/output tokens; GPT-5 runs $1.25/$10.
- Which has the larger context window?
- Gemini 3.1 Pro, at 1,000,000 tokens versus 400,000.
- How accurate are these token counts?
- Gemini 3.1 Pro: Approximated with o200k_base; drift typically ~3% on English and code. GPT-5: Exact tokenization via the canonical OpenAI vocab (o200k_base). The dollar math itself is exact once the token count is known.
Both prices are computed from tokenmath's verified pricing table. Rates sourced from ai.google.dev and openai.com, verified 2026-07-06. Vendor pricing changes often — confirm before you commit.