Side by side
See the difference.
Scores, capabilities and costs for the models you’re considering.
Scroll to compare models →
| 3 models | googleGoogle: Gemini 3.8 FlashRemove × | kreaKrea: K3 EditRemove × | openaiOpenAI: GPT-6.1 SolRemove × |
|---|---|---|---|
| What’s different | Gemini 3.8 Flash is Google's most intelligent Flash model with significant gains from 3.7 Flash across software engineering, agentic tasks, and multi-step reasoning. | Edits one to eight reference images from a text prompt, returning one image. | Requires Responses API for tools; offers Ultrafast inference with US or EU data residency. |
| Date | Sep 2, 2026Listed | Oct 15, 2026Listed | Sep 29, 2026Released |
API prices USD · provider configuration shown below | |||
| Listed rate | $0.375 in / $1.875 out/ 1M tokensGoogle AI Studio · via OpenRouter · all routes ↗google-ai-studio/flex · tier=flex | See provider quotes | $1.00 in / $5.00 out/ 1M tokensOpenAI · via OpenRouter · all routes ↗openai/flex · tier=flex · prompt ≤ 272000 tokens |
| Provider endpoints | 9 endpoints → | 0 endpoints → | 10 endpoints → |
Benchmarks Each result retains its evaluation configuration | |||
| AA Intelligence ↗4.3.2 | Not reported | Not reported | 51.83 ↗Effort: max · MaxIndependent result |
| AA-Briefcase ↗1.1 | Not reported | Not reported | 1,556.34 ↗Effort: max · MaxIndependent result |
| GDPval-AA ↗2.1 | Not reported | Not reported | 1,592.1 ↗Effort: max · MaxIndependent result |
| AutomationBench-AA ↗AA Index 4.3.2 collection | Not reported | Not reported | 64.87% ↗Effort: max · MaxIndependent result |
| Terminal-Bench 4.0 · AA ↗ | Not reported | Not reported | 56.06% ↗Effort: max · MaxHarness unreportedIndependent result |
| SciCode ↗AA Index 4.3.2 collection | Not reported | Not reported | 54.17% ↗AA · Under reviewEffort: max · MaxIndependent result |
| Humanity’s Last Exam ↗AA Index 4.3.2 collection | Not reported | Not reported | 52.92% ↗Effort: max · MaxIndependent result |
| GDP.pdf ↗AA Index 4.3.2 collection | Not reported | Not reported | 31% ↗Effort: max · MaxIndependent result |
| CritPt ↗AA Index 4.3.2 collection | Not reported | Not reported | 31.71% ↗AA · Under reviewEffort: max · MaxIndependent result |
| AA-Omniscience ↗AA Index 4.3.2 collection | Not reported | Not reported | 41.53 ↗Effort: max · MaxIndependent result |
| AA-LCR ↗1.1 | Not reported | Not reported | 83% ↗Effort: max · MaxIndependent result |
| LiveBench ↗2.0 · 2026-10-14 (dataset 2026-09-02) | Not reported | Not reported | 85.41% ↗Effort: max · single run · equal mean of 7 suitesIndependent result |
| LiveBench Design ↗2.0 · 2026-10-14 (dataset 2026-09-02) | Not reported | Not reported | 82.1% ↗Effort: max · single runIndependent result |
| LiveBench Research ↗2.0 · 2026-10-14 (dataset 2026-09-02) | Not reported | Not reported | 83.5% ↗Effort: max · single runIndependent result |
| LiveBench Browser Use ↗2.0 · 2026-10-14 (dataset 2026-09-02) | Not reported | Not reported | 88% ↗Effort: max · single runIndependent result |
| LiveBench Coding ↗2.0 · 2026-10-14 (dataset 2026-09-02) | Not reported | Not reported | 96% ↗Effort: max · single runIndependent result |
| LiveBench Agentic ↗2.0 · 2026-10-14 (dataset 2026-09-02) | Not reported | Not reported | 73% ↗Effort: max · single runIndependent result |
| LiveBench Data Analysis ↗2.0 · 2026-10-14 (dataset 2026-09-02) | Not reported | Not reported | 82.7% ↗Effort: max · single runIndependent result |
| LiveBench Reasoning ↗2.0 · 2026-10-14 (dataset 2026-09-02) | Not reported | Not reported | 92.6% ↗Effort: max · single runIndependent result |
Specifications | |||
| Context window | 1M | 65.5K | 1.1M |
| Max output | 65.5K | 59K | 128K |
| Parameters | — | — | — |
| Tool calling | Supported | Not reported | Supported |
| Structured output | Supported | Not reported | Supported |
| Image input | Supported | Supported | Supported |
| Audio input | Supported | Not reported | Not reported |
| Video input | Supported | Not reported | Not reported |
| Reasoning | Supported | Not reported | Supported |
| Prompt caching | Supported | Not reported | Supported |