# Gemini 3.8 Flash vs GPT-5.6 Luna

As of October 21, 2026, Gemini 3.8 Flash costs 275% more per input token than GPT-5.6 Luna ($0.75 vs $0.20 per 1M) and has a 1M-token context window versus 922K.

- Page: https://models.litellm.ai/compare/gemini-3.8-flash-vs-gpt-5.6-luna
- Gemini 3.8 Flash: https://models.litellm.ai/models/gemini-3.8-flash
- GPT-5.6 Luna: https://models.litellm.ai/models/gpt-5.6-luna

## Side by side

| | Gemini 3.8 Flash | GPT-5.6 Luna |
|---|---|---|
| Input, per 1M tokens | $0.75 | $0.20 |
| Output, per 1M tokens | $3.75 | $1.20 |
| Cache read, per 1M tokens | $0.075 | $0.02 |
| Cache write, per 1M tokens | – | $0.25 |
| Context window | 1,048,576 tokens | 922,000 tokens |
| Max output | 65,536 tokens | 128,000 tokens |
| Image input | Yes | Yes |
| Tool calling | Yes | Yes |
| Structured output (JSON) | Yes | Yes |
| Prompt caching | Yes | Yes |
| Reasoning | Yes | Yes |
| Providers on LiteLLM | 5 | 9 |
| Price shown is from | Google AI Studio | OpenAI |
| Added to LiteLLM | September 2, 2026 | July 9, 2026 |

Headline prices in USD per 1M tokens, from LiteLLM's model price file.

## Cheapest provider for each

- Gemini 3.8 Flash: Google AI Studio at $0.75 input per 1M tokens (`gemini/gemini-3.8-flash`)
- GPT-5.6 Luna: OpenAI at $0.20 input per 1M tokens (`gpt-5.6-luna`)

## Benchmark scores

| Benchmark | Gemini 3.8 Flash | GPT-5.6 Luna |
|---|---|---|
| Epoch Capabilities Index (Overall) | 157.4 | 156.5 |
| FrontierCode (Coding and agents) | 41.2% (medium) | 39.8% |
| APEX-Agents (Coding and agents) | 64.3% | 43.0% |
| GPQA Diamond (Reasoning and knowledge) | 95.4% (high) | 91.6% (max) |
| ARC-AGI-2 (Reasoning and knowledge) | 89.2% (high) | 59.5% (max) |
| SimpleQA Verified (Reasoning and knowledge) | 69.7% (high) | 41.0% (max) |
| FrontierMath (tiers 1-3) (Math) | 68.4% (high) | 82.1% (max) |
| FrontierMath (tier 4) (Math) | 22.0% (high) | 61.0% (max) |
| OTIS Mock AIME 2024-2025 (Math) | 98.9% (high) | 98.3% (max) |

From the Epoch AI Benchmarking Hub (https://epoch.ai/benchmarks), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/), data as of September 2, 2026: Epoch's own runs and results it collects from each benchmark's leaderboard. Higher is better.

## When to pick which

- Gemini 3.8 Flash: Larger context window: 1,048,576 vs 922,000 tokens.
- GPT-5.6 Luna: Costs 73% less per input token: $0.20 vs $0.75 per 1M.
- GPT-5.6 Luna: Returns up to 128,000 tokens per response, vs 65,536.

## FAQ

### Which is cheaper, Gemini 3.8 Flash or GPT-5.6 Luna?

GPT-5.6 Luna is cheaper at 3 input tokens per output token: $0.45 per 1M tokens blended ($0.20 input, $1.20 output), versus $1.50 ($0.75 and $3.75) for Gemini 3.8 Flash, as of October 21, 2026.

### Which has the larger context window, Gemini 3.8 Flash or GPT-5.6 Luna?

Gemini 3.8 Flash has the larger context window: 1,048,576 tokens, versus 922,000 for GPT-5.6 Luna.

### Can I switch from Gemini 3.8 Flash to GPT-5.6 Luna without changing my code?

Yes. Through LiteLLM both use the OpenAI format, so switching means changing the model name from "gemini/gemini-3.8-flash" to "gpt-5.6-luna", or pointing a LiteLLM gateway model alias at the other model. You can also use one as a fallback for the other.