Gemini 3.6 Flash vs GPT-5.6 Luna: which should you choose in 2026?
Gemini 3.6 Flash (by Google) and GPT-5.6 Luna (by OpenAI) are compared below. Here is how they stack up on benchmarks, price, and capabilities, and which one to pick in 2026.
No protocol-matched benchmark supports a controlled winner. Looking only at published point scores, Gemini 3.6 Flash is higher on 3 benchmarks (AutomationBench-AA, GPQA Diamond, Humanity's Last Exam), while GPT-5.6 Luna is higher on 0 benchmarks.
The evaluation setups are not fully aligned, so these are not controlled head-to-head wins.
GPT-5.6 Luna is about 13.3× cheaper on a blended 3:1 input/output basis ($0.23 vs $3.00 per 1M tokens).
Choose Gemini 3.6 Flash if…
- • your own prompt tests favor its output; shared comparable evidence does not identify a unique advantage
Choose GPT-5.6 Luna if…
- • lower blended API cost matters for your workload
The benchmark count includes only results measured with a matching benchmark version and protocol. Arena scores are shown separately. Missing, preliminary, and incompatible data is not treated as a controlled win. Published point-score comparisons are labeled separately when protocol details are incomplete. For text-output models, the verdict also compares token pricing and context windows.
Performance benchmarks
Every value links to its source. A dash means that no reviewed result is available for that exact model and protocol.
| Benchmark | Gemini 3.6 Flash | GPT-5.6 Luna |
|---|---|---|
A 657-task benchmark of multi-step work across simulated SaaS applications in six business domains. The Artificial Analysis protocol reports a guardrail-aware score and is distinct from both the public Zapier split and the unrelated dynamic AutoBench framework. Guardrail-aware score Protocols differ or are incompletely disclosed; not counted as a controlled win | 51.1% Artificial Analysis | 42.2% Artificial Analysis |
A benchmark of difficult, verifiable information-seeking questions designed to measure an agent's ability to locate hard-to-find facts through web browsing. Accuracy | — | — |
A verified computer-use benchmark in which multimodal agents operate desktop applications and are graded from the resulting environment state. Mean task reward | — | — |
A long-horizon software-engineering benchmark with 113 original tasks graded by hand-written tests. Pass@1 Protocols differ or are incompletely disclosed; not counted as a controlled win | 49.0%±5.0 DataCurve | 67.0%±4.0 DataCurve |
A human-validated subset of real GitHub issues used to measure whether a coding agent can produce repository patches that resolve the associated tests. Resolved | — | — |
Version 2.1 of the benchmark for completing realistic tasks in terminal environments. Harness, resource limits, and attempt count are part of the protocol. Mean task success | — | 75.7%±1.3 Terminal-Bench |
The highest-quality subset of Graduate-Level Google-Proof Q&A, designed to test expert-level scientific reasoning in biology, physics, and chemistry. Accuracy Protocols differ or are incompletely disclosed; not counted as a controlled win | 92.8% Artificial Analysis | 91.1% Artificial Analysis |
A 2,500-question expert-level benchmark spanning dozens of academic fields. Tool-assisted and no-tools results are separate protocols and must not be merged. Accuracy Protocols differ or are incompletely disclosed; not counted as a controlled win | 40.8% Artificial Analysis | 39.5% Artificial Analysis |
A contamination-resistant benchmark refreshed on a fixed release cadence. Scores from different LiveBench releases must never be compared as the same protocol. Mean of category averages Protocols differ or are incompletely disclosed; not counted as a controlled win | 73.6% LiveBench | 73.6% LiveBench |
Benchmark sources
- AutomationBench-AA: Agentic SaaS Workflow Benchmark
- DeepSWE 1.1 Leaderboard
- terminal-bench@2.1 Leaderboard
- GPQA Diamond Benchmark Leaderboard
- Humanity's Last Exam Benchmark Leaderboard
- LiveBench 2026-06-25 Leaderboard
Reviewed evidence last updated 2026-08-09.
Pricing, capabilities, and model facts
| Feature | Gemini 3.6 Flash | GPT-5.6 Luna |
|---|---|---|
| Context & model facts | ||
| Developer | OpenAI | |
| API provider | OpenAI | |
| Input context | 1,048,576 tokens | 1,050,000 tokens |
| Maximum output | 65,536 tokens | 128,000 tokens |
| Released | — | — |
| Added to Writingmate | Jul 21, 2026 | Jul 9, 2026 |
| License | — | — |
| Knowledge cutoff | — | 2026-02-16 |
| Capabilities | ||
| Inputs | Text, Image, Video, File, Audio | File, Image, Text |
| Outputs | Text | Text |
| Tool use | Yes | Yes |
| Reasoning | Yes | Yes |
| Vision | Yes | Yes |
| Image Generation | No | No |
| Video Generation | No | No |
| API pricing | ||
| Input (per 1M tokens) | $1.50 | $0.10 |
| Output (per 1M tokens) | $7.50 | $0.60 |
| Blended 3:1 input/output | $3.00 | $0.22 |
| API performance | ||
| p95 latency | Not measured | Not measured |
| Output throughput | Not measured | Not measured |
| Writingmate shows API performance only when both models have enough observations from the same measurement window, prompt profile, and provider. Third-party latency values are not copied into this table. | ||
Data sources
- Writingmate model catalog (pricing, limits, and availability)
Catalog data last updated Jul 21, 2026.
Why Pay for Multiple Subscriptions?
Comparing Gemini 3.6 Flash from Google with GPT-5.6 Luna from OpenAI? Instead of managing separate API keys and subscriptions, get both with Writingmate.
| Plan | Price | Gemini 3.6 Flash | GPT-5.6 Luna | AI Images | AI Video |
|---|---|---|---|---|---|
Writingmate Pro Most popular | $20/mo | Included | Included | Nano Banana Pro, FLUX.2, DALL-E & more | Sora 2, VEO 3.1 |
Writingmate Ultimate Power users | $60/mo | Included | Included | Nano Banana Pro, FLUX.2, DALL-E & more | Sora 2, VEO 3.1 |
Gemini 3.6 Flash vs GPT-5.6 Luna FAQ
Which is better, Gemini 3.6 Flash or GPT-5.6 Luna?
No protocol-matched benchmark supports a controlled winner. Looking only at published point scores, Gemini 3.6 Flash is higher on 3 benchmarks (AutomationBench-AA, GPQA Diamond, Humanity's Last Exam), while GPT-5.6 Luna is higher on 0 benchmarks. The evaluation setups are not fully aligned, so these are not controlled head-to-head wins. GPT-5.6 Luna is about 13.3× cheaper on a blended 3:1 input/output basis ($0.23 vs $3.00 per 1M tokens).
Which model is cheaper to use through an API?
GPT-5.6 Luna is about 13.3× cheaper on a blended 3:1 input/output basis ($0.23 vs $3.00 per 1M tokens).
Which model supports more context?
The published context windows are close enough that this page does not treat the difference as a material advantage.
Can I switch between Gemini 3.6 Flash and GPT-5.6 Luna?
Yes. Use the model selector on this page to open any current Writingmate model comparison. You can also run the same prompt with both models in Writingmate.