Gemini 3.5 Flash
Gemini 3.5 Flash is an AI model released by Google on Tuesday, May 19 2026, 47 days after Gemma 4. Benchmark results (shown below) cover Toolathlon, Finance Agent v2, MMMU-Pro, BullshitBench v2, Gray Swan IPI, SWE-Bench Pro, and 17 more.
API pricing
| USD per 1M tokensEvery price here is for one million tokens. A token is roughly three-quarters of a word, so a million tokens is about 750,000 words of text. | |
|---|---|
| InputWhat you pay for everything you send the model — your question, plus any documents or earlier conversation you include with it. | $1.50 |
| Cached inputA reduced rate for text you send over and over. If every request starts with the same instructions or the same document, the provider keeps a copy ready and charges less to read it again. | $0.15 |
| OutputWhat you pay for the text the model writes back. It is normally the dearer half: producing an answer costs more than reading one. | $9.00 |
- Output prices include reasoning tokens.
SourceVerified August 18, 2026
Benchmarks
ToolathlonGeneral tool use — Tests how well the AI uses everyday real-world tools and apps to get things done. Higher is better.
56.5%
#1Best published Toolathlon score of all tracked models
Finance Agent v2Agentic financial analysis — Tests the AI on real financial-analysis work, like digging through reports and making sound decisions. Higher is better.
57.9%
#1Best published Finance score of all tracked models
MMMU-ProMultimodal reasoning — A tougher version of MMMU — college-level questions that mix images, diagrams, and text together. Higher is better.
83.6%
#1Best published MMMU-Pro score of all tracked models
BullshitBench v2Nonsense detection — Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better.
20%
#46 of 67Best: Claude Opus 4.8 · 95%
Gray Swan IPIPrompt injection robustness — Attackers hide malicious instructions inside content the AI reads — a web page, an email, a document — and try to hijack what it does. Gray Swan's indirect prompt injection benchmark measures how often such an attack succeeds when the attacker gets a single try. Lower is better.
14.1%
k = 1
54.2%
k = 10
60.5%
k = 15
k = 1
#12 of 13Best: Claude Opus 5 · 0.2%
k = 10
#12 of 13Best: Claude Opus 5 · 1.6%
k = 15
#12 of 13Best: Claude Opus 5 · 2%
SWE-Bench ProAgentic coding — Can the AI fix real bugs in real software? It's handed actual problems from open-source projects and has to write code that genuinely solves them. Higher is better.
55.1%
#12 of 20Best: Claude Fable 5 · 80.3%
CursorBench v3.2Agentic coding — Cursor's own test of harder, real-world coding tasks inside a code editor, on the refreshed v3.2 task set. Scores aren't comparable with v3.1. Higher is better.
48.8%
#15 of 15Best: Claude Fable 5 · 70.5%
CursorBench v3.1Agentic coding — Cursor's own test of harder, real-world coding tasks inside a code editor. Higher is better.
49.8%
#9 of 12Best: Claude Fable 5 · 72.9%
DeepSWE 1.1Agentic coding — Artificial Analysis' independent test of deep, agentic software-engineering work — the AI has to plan and carry out substantial coding tasks end to end. (Version 1.1 of the test.) Higher is better.
37%
#14 of 18Best: GPT-5.6 Sol · 73%
MLE-BenchML engineering — Can the AI do the work of a machine-learning engineer? It competes in real Kaggle competitions — building, training, and tuning models end to end — and the score reflects how well it places. Higher is better.
49.7%
#2 of 3Best: Gemini 3.6 Flash · 63.9%
Terminal-Bench 2.1Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Higher is better.
76.2%
#16 of 25Best: GPT-5.6 Sol · 88.8%
MCP AtlasMulti-step tool use — Can the AI chain together many tools and steps to complete one bigger task, rather than doing just a single thing? Higher is better.
83.6%
#3 of 11Best: Muse Spark 1.1 · 88.1%
BU BenchBrowser agent — Can the AI drive a real web browser to finish tasks — clicking, filling forms, and navigating sites the way a person would? Run by Browser Use on their BU Bench task set. Higher is better.
58%
#5 of 7Best: Claude Opus 4.8 · 74%
Humanity's Last ExamMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects, written so you can't just look up the answer. “No tools” means the AI answers on its own. Higher is better.
40.2%
no tools
#10 of 19Best: Claude Opus 5 · 56.3%
ARC-AGI-2Abstract reasoning — Puzzle-style tests of abstract reasoning and pattern-finding — the kind of thing people find easy but AIs often struggle with. Higher is better.
72.1%
#5 of 11Best: GPT-5.5 · 84.6%
OSWorld-VerifiedAgentic computer use — Can the AI actually operate a computer — clicking, typing, and using real apps — to finish tasks on its own? Higher is better.
78.4%
#8 of 17Best: Qwen3.8-Max · 86.1%
GDPval-AAKnowledge work — Measures how well the AI does economically valuable knowledge work, judged against human experts. Shown as a rating (like a chess Elo) — higher is better.
1656
#6 of 9Best: Claude Fable 5 · 1932
GDPval-AA v2Knowledge work — economically valuable knowledge work (v2, re-based Elo)
1349
#13 of 16Best: Claude Opus 5 · 1861
CharXiv ReasoningChart reasoning — Can the AI read and reason about complex charts and figures, not just text? Higher is better.
84.2%
#6 of 14Best: Muse Spark · 88.9%
Blueprint-Bench 2Spatial reasoning — Can the AI reason about space and layout — for example, understanding a floor plan or blueprint? Higher is better.
33.6%
#2 of 6Best: GPT-5.5 · 36.2%
MRCR v2 (8-needle)Long context — Tests whether the AI can find specific details buried inside a very long document (around 128k tokens — roughly a long book). Higher is better.
77.3%
128k average
26.6%
1M pointwise
128k average
#5 of 7Best: Gemini 3.7 Flash · 97%
1M pointwise
#2 of 4Best: Gemini 3.7 Flash · 62.5%
Arena Elo (Text)Community preference — Real people chat with two anonymous AIs side by side and vote for the answer they prefer. Votes become a chess-style Elo rating on arena.ai — it measures which AI people actually like, not test scores. Higher is better.
1476
#17 of 32Best: Claude Fable 5 · 1509
Arena Elo (Code)Community preference (code) — Like the text arena, but people vote on which AI writes better code. The votes become a chess-style Elo rating on arena.ai. Higher is better.
1499
#24 of 50Best: Kimi K3 · 1679
Gemini 3.5 Flash — frequently asked questions
- When was Gemini 3.5 Flash released?
- Gemini 3.5 Flash was released by Google on Tuesday, May 19 2026.
- Who made Gemini 3.5 Flash?
- Gemini 3.5 Flash was built by Google. Builds the Gemini family of models through Google DeepMind. Integrates AI across Google products.
- How much does Gemini 3.5 Flash cost?
- Gemini 3.5 Flash costs $1.50 per million input tokens and $9.00 per million output tokens through the Google API. Cached input is $0.15 per million tokens. Output prices include reasoning tokens. Rates are pay-as-you-go API prices verified against Google's published pricing on August 18, 2026.
- What benchmark scores did Gemini 3.5 Flash get?
- Gemini 3.5 Flash reports 26 tracked benchmark scores — BullshitBench v2: 20%; Gray Swan IPI (k = 1): 14.1%; Gray Swan IPI (k = 10): 54.2%; Gray Swan IPI (k = 15): 60.5%; SWE-Bench Pro: 55.1%; CursorBench v3.2: 48.8%; CursorBench v3.1: 49.8%; DeepSWE 1.1: 37%; MLE-Bench: 49.7%; Terminal-Bench 2.1: 76.2%; MCP Atlas: 83.6%; Toolathlon: 56.5%; BU Bench: 58%; Humanity's Last Exam (no tools): 40.2%; ARC-AGI-2: 72.1%; OSWorld-Verified: 78.4%; Finance Agent v2: 57.9%; GDPval-AA: 1656; GDPval-AA v2: 1349; CharXiv Reasoning: 84.2%; MMMU-Pro: 83.6%; Blueprint-Bench 2: 33.6%; MRCR v2 (8-needle) (128k average): 77.3%; MRCR v2 (8-needle) (1M pointwise): 26.6%; Arena Elo (Text): 1476; Arena Elo (Code): 1499. Scores are the figures published at release by Google. It holds the best score among all models tracked here on Toolathlon, Finance Agent v2 and MMMU-Pro.
- Is Gemini 3.5 Flash open source?
- No. Gemini 3.5 Flash is a proprietary model. The weights are not published — it is available only through the provider's own API, apps, or partner platforms.
- What came before and after Gemini 3.5 Flash?
- Google's previous tracked release was Gemma 4 on Apr 2 2026, 47 days earlier. It was followed by Gemini 3.6 Flash on Jul 21 2026.
Compare Gemini 3.5 Flash with
Gemini 3.5 FlashvsGemini 3.7 FlashGemini 3.5 FlashvsGPT-5.6-CyberGemini 3.5 FlashvsClaude Opus 5Gemini 3.5 FlashvsMuse GlimmerGemini 3.5 FlashvsGrok 4.6Gemini 3.5 FlashvsDeepSeek-V4-Pro-0813Gemini 3.5 FlashvsMistral Medium 3.5Gemini 3.5 FlashvsKimi K3Gemini 3.5 FlashvsGLM-5.3Gemini 3.5 FlashvsQwen3.8-27B