Models › Gemini 3.6 Flash
Google DeepMind • Launched July 2026 as the default Flash model
Gemini 3.6 Flash is Google’s efficient workhorse for everyday work
Starting provider API rate; context tiers apply
See what 3.6 Flash does best, then review its limits, benchmarks, and pricing below
Gemini 3.6 Flash is Google’s efficient workhorse for everyday work
Every 3.6 Flash result below comes from a named source and is labeled by protocol, so each score reflects how it was measured
| Benchmark | Score | Evaluation details | Source |
|---|---|---|---|
GPQA Diamond The highest-quality subset of Graduate-Level Google-Proof Q&A, designed to test expert-level scientific reasoning in biology, physics, and chemistry. |
92.8% Accuracy |
Version: GPQA Diamond 198 · Split: 198 Diamond questions · Harness: Artificial Analysis independent evaluation · Reasoning effort: high | Methodology Artificial Analysis Independently reproduced |
Humanity's Last Exam A 2,500-question expert-level benchmark spanning dozens of academic fields. Tool-assisted and no-tools results are separate protocols. |
38% Accuracy |
Version: Humanity's Last Exam · Protocol: No external tools · Harness: Artificial Analysis independent evaluation · Reasoning effort: high | Methodology Artificial Analysis Independently reproduced |
GDPval-AA v2 A knowledge-work benchmark scoring long, realistic professional tasks against expert-produced deliverables across dozens of occupations. |
1,421 Elo score |
Version: GDPval-AA v2 · Harness: Artificial Analysis independent evaluation · Scoring: Pairwise Elo comparison against expert-produced work · Reasoning effort: high | Methodology Artificial Analysis Independently reproduced |
SWE-bench Pro An extended software-engineering benchmark that tests whether a model can resolve harder, more realistic repository issues than earlier SWE-bench sets. |
58.7% Pass@1 |
Version: SWE-bench Pro · Harness: Google DeepMind evaluation suite · Reasoning effort: high | Methodology Google DeepMind Provider-reported |
OSWorld-Verified A computer-use benchmark that scores whether an agent can complete real desktop tasks inside a verified operating-system environment. |
83.0% Task success rate |
Version: OSWorld-Verified · Harness: Google DeepMind evaluation suite · Reasoning effort: high | Methodology Google DeepMind Provider-reported |
Arena rankings capture how people vote between anonymous responses. 3.6 Flash’s human preference standing is reported per leaderboard, since results can vary widely by task
3.6 Flash debuted at #12 on Arena’s Text leaderboard with 1,485 points, landing in the top 20 at launch
Gemini 3.6 Flash runs a 1,000,000-token context window with image, audio, and video input, adjustable reasoning, and pricing detailed below
| Model | |
| Provider | Google DeepMind |
| Release date | Jul 21, 2026 |
| Catalog added | Sep 22, 2026 |
| Context window | 1,000,000 tokens |
| Maximum output | 64,000 tokens |
| Knowledge cutoff | March 2026 |
| License | Proprietary |
| Capabilities and API | |
| Input types | Image, Audio, Video, Text |
| Output types | Text |
| Reasoning | Supported (thinking mode, effort-adjustable) |
| Provider endpoint tool parameters | Accepts tool parameters |
| Base provider API input | $0.75 / 1M tokens |
| Base provider cached input | $0.15 / 1M tokens |
| Base provider API output | $3.75 / 1M tokens |
Provider token rates are context-dependent. The base rates apply below the listed prompt thresholds; the matching tier applies at or above each threshold. These are reference provider API rates, not Pyxa AI checkout charges — access in Pyxa AI follows the allowances of your Pyxa AI plan.
Each 3.6 Flash benchmark and specification links to its original source, so you can check where every figure comes from
Open another model’s detail page or compare it directly with Gemini 3.6 Flash. Suggestions come from the current catalog
Put Gemini 3.6 Flash to work on your coding, research, and long-document tasks inside Pyxa AI