Gemini 3.8 Flash vs GPT-6 Luna: cost per task
GPT-6 Luna is cheaper on 14 of 14 tasks and about 16.4x cheaper on average (geometric mean). Gemini 3.8 Flash has the higher LMArena Text score (1496 vs 1443).
| Task | Gemini 3.8 Flash | GPT-6 Luna | Difference |
|---|---|---|---|
| Answer a chat question | 0.68¢ | 0.06¢ | 11.4x |
| Reply to a customer email | 0.25¢ | <0.01¢ | 34.3x |
| Write a 1,500-word blog post | 1.5¢ | 0.12¢ | 12.5x |
| Write 10 product descriptions | 1.2¢ | 0.05¢ | 23.0x |
| Translate 1,000 words | 1.9¢ | 0.08¢ | 23.5x |
| Proofread one page (500 words) | 0.51¢ | 0.04¢ | 11.4x |
| Summarize a 20-page PDF | 1.6¢ | 0.16¢ | 10.0x |
| Extract data from an invoice | 0.65¢ | 0.03¢ | 20.4x |
| Analyze a 200-row spreadsheet | 3.9¢ | 0.19¢ | 20.6x |
| Answer from 5 retrieved documents (RAG) | 1.2¢ | 0.05¢ | 22.2x |
| Classify 100 support tickets | 2.9¢ | 0.06¢ | 44.3x |
| Review 500 lines of code | 1.3¢ | 0.31¢ | 4.2x |
| Write a SQL query | 0.67¢ | 0.04¢ | 16.5x |
| Run a 50-step coding agent | 50¢ | 4.1¢ | 12.2x |
Specs and list prices
| Gemini 3.8 Flash | GPT-6 Luna | |
|---|---|---|
| Input per 1M | $0.82 | $0.10 |
| Output per 1M | $4.12 | $0.50 |
| Cached input per 1M | $0.083 | $0.01 |
| Context | 1M | 1.05M |
| Providers | 7 | 7 |
| Reasoning | always | default |
| LMArena Text | 1496 | 1443 |
| Released | Sep 2, 2026 | Sep 22, 2026 |
FAQ
Is Gemini 3.8 Flash or GPT-6 Luna cheaper?
GPT-6 Luna is cheaper on 14 of 14 tasks and about 16.4x cheaper on average (geometric mean). Gemini 3.8 Flash has the higher LMArena Text score (1496 vs 1443).
How much does Gemini 3.8 Flash cost compared to GPT-6 Luna per token?
Gemini 3.8 Flash: $0.82 input and $4.12 output per 1M tokens. GPT-6 Luna: $0.10 input and $0.50 output. Token prices alone mislead - models use different numbers of tokens for the same task, which is what the task table shows.