GPT-6 Luna vs Gemini 3.8 Flash
Short answer
GPT-6 Luna is about 8x cheaper than Gemini 3.8 Flash for a typical 3:1 input-to-output token mix ($0.20 vs $1.50 per 1M tokens). Gemini 3.8 Flash scores higher on ARC-AGI-2 (89.2% vs 59.3%).
Prices updated Nov 13, 2026 (APIpulse, CC BY 4.0) · Benchmarks: ARC Prize, Vals.ai · all models
| GPT-6 Lunacheaper | Gemini 3.8 Flash | |
|---|---|---|
| Provider | OpenAI | |
| Input / 1M tokens | $0.10 | $0.75 |
| Output / 1M tokens | $0.50 | $3.75 |
| Blended (3:1) / 1M | $0.20 | $1.50 |
| Context window | 1.05M | 1M |
| ARC-AGI-2 | 59.3% (Max) | 89.2% (High) |
| SWE-bench Verified | — | 80% |
API cost calculator
Enter your monthly usage in millions of tokens (1M tokens ≈ 750,000 words).
Standard rates; batch processing and prompt caching can lower the bill.
Frequently asked questions
Which is cheaper, GPT-6 Luna or Gemini 3.8 Flash?
GPT-6 Luna is about 8x cheaper than Gemini 3.8 Flash for a typical 3:1 input-to-output token mix ($0.20 vs $1.50 per 1M tokens).
How much do GPT-6 Luna and Gemini 3.8 Flash cost per month?
For 10M input and 2M output tokens a month: GPT-6 Luna costs $2 and Gemini 3.8 Flash costs $15.
More: GPT-6 Luna pricing · Gemini 3.8 Flash pricing · all API prices
Benchmarks updated Nov 15, 2026.