Gemini 3.1 Pro vs Gemini 2.5 Flash
Side-by-side comparison of API pricing, specs, benchmarks, and capabilities
| Specification |
by Google
|
by Google
|
|---|---|---|
| Overview | ||
| Status | Current flagship | Deprecated budget |
| Released | Feb 19, 2026 | Jun 17, 2025 |
| Pricing per million tokens | ||
| Input | $2.00/Mtok | $0.300/Mtok |
| Output | $12.00/Mtok | $2.50/Mtok |
| Blended avg | $4.50/Mtok | $0.850/Mtok |
| Cached input | $0.200/Mtok | — |
| Specifications | ||
| Context window | 2M tokens | 1M tokens |
| Parameters | Proprietary | Proprietary |
| Speed (TPS) | — | — |
| Modalities | ||
| Input |
textimageaudiovideo
|
textimageaudiovideo
|
| Benchmarks source: Vellum LLM Leaderboard | ||
| Avg benchmark score | 74.8 | 50.2 |
| Perf / dollar | 16.6 | 59.1 |
| GPQA Diamond | 94.3 | 78.3 |
| Terminal-Bench 2.1 | 70.3 | 16.9 |
| Humanity's Last Exam | 44.4 | 12.1 |
| ARC-AGI 2 | 77.1 | 15 |
| SWE-Bench Verified | — | 38 |
| LiveCodeBench | — | 63.5 |
| MCP Atlas | 69.2 | — |
| BrowseComp | 85.9 | — |
| OSWorld-Verified | 76.2 | — |
| AIME 2025 | — | 78 |
| MMMLU | — | 78 |
| HumanEval | 80.6 | — |
| MATH 500 | — | 72 |
| Independent composite scores each on its own scale — not part of the average above | ||
| LMSYS Chatbot Arena | 1486.5 ELO | 1410.2 ELO |
| AA Intelligence Index | 47.7 | — |
| AA Agentic Index | 23 | — |
| AA-Omniscience Index (−100–100) | 31.9 | — |
| Providers | ||
| Available from |
Google — $2.00/$12.00/Mtok
|
Google — $0.300/$2.50/Mtok
|
Cost at scale
1M tokens · 50/50 input/output| Volume | Gemini 3.1 Pro | Gemini 2.5 Flash | Savings |
|---|---|---|---|
| 1M tokens | $4.5 | $0.85 | $3.65 (81.1%) |
| 10M tokens | $45 | $8.5 | $36.5 (81.1%) |
| 100M tokens | $450 | $85 | $365 (81.1%) |
| 1000M tokens | $4500 | $850 | $3650 (81.1%) |
When to pick which
distilled from the pricing and spec data above- your workload re-reads context (agents, RAG, long chats): cached input costs $0.200/Mtok — 90% off its list input price, a discount Gemini 2.5 Flash doesn't offer
- you need the longer context: 2M tokens vs 1M (2×)
- raw capability matters more than price: 74.8 vs 50.2 average benchmark score
- cost dominates: $0.850/Mtok blended vs $4.50 — 81.1% less on the same 50/50 token mix
Price movement
recorded changes since we began tracking each model · last checked Aug 15, 2026| Model | Changes | Latest move | Old → new $/M |
|---|---|---|---|
| Gemini 3.1 Pro | 0 | — no move | Stable since we began tracking — no change recorded |
| Gemini 2.5 Flash | 2 | price rise on Jul 4, 2026 | $0.075/$0.3 → $0.3/$2.5/Mtok |
The cheaper option here — Gemini 2.5 Flash costs $0.850/Mtok blended on Google.
Summary
Gemini 3.1 Pro by Google costs $2.00/Mtok input and $12.00/Mtok output, with a 2M-token context window. It supports text, image, audio, video input.
Gemini 2.5 Flash by Google costs $0.300/Mtok input and $2.50/Mtok output, with a 1M-token context window. It supports text, image, audio, video input.
On a blended cost basis, Gemini 2.5 Flash is 81.1% cheaper than Gemini 3.1 Pro.
On benchmarks, Gemini 3.1 Pro scores higher (74.8 vs 50.2) on average. In terms of value, Gemini 2.5 Flash has better performance per dollar (59.1 vs 16.6).
Note: Pricing is per million tokens. Actual costs vary with usage patterns, prompt caching, and batch discounts. Always verify against official provider pricing pages.