google/gemini-3.8-flash

模型发布 OpenRouter 模型 原文语言:中文

OpenRouter 于 2026-09-02 收录该模型(收录时间不等于模型正式发布日期);上下文:1,048,576 tokens;输入价:$0.8250/百万 tokens;输出价:$4.1250/百万 tokens;定位(来源描述):Gemini 3.8 Flash is Google's most intelligent Flash model with significant gains from 3.7 Flash…

全文

Google: Gemini 3.8 Flash

google/gemini-3.8-flash

Compare

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.

Modalities

In / out price

50% off

$0.75 / $3.75per 1M

Context

1.0M

Released

Sep 2, 2026

Compare

Providers

Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick: Balanced (price + speed), Nitro (fastest), or Floor (cheapest).

Latency / throughput

50% off

$1.50$0.750$7.50$3.750$0.15$0.0750$0.15$0.07502.64s76 tok/s

100.00%

50% off

$1.50$0.750$7.50$3.750$0.15$0.0750$0.15$0.07502.70s84 tok/s

100.00%

Not used in Standard routing:Why these endpoints are not used

Flex50% off

$0.75$0.375$3.75$1.875$0.075$0.0375$0.075$0.03752.44s76 tok/s

100.00%

Flex50% off

$0.75$0.375$3.75$1.875$0.075$0.0375$0.075$0.037515.45s82 tok/s

93.85%

50% off

$1.65$0.825$8.25$4.125$0.165$0.0825$0.165$0.08253.96s111 tok/s

99.97%

Priority50% off

$2.70$1.350$13.50$6.750$0.27$0.1350$0.27$0.13502.10s67 tok/s

99.71%

Priority50% off

$2.70$1.350$13.50$6.750$0.27$0.1350$0.27$0.13501.83s34 tok/s

99.77%

Provider50% offInput / Output /M

Input $0.75 / Output $3.7550% off

76 tok/s100.00% uptime

Input $0.75 / Output $3.7550% off

84 tok/s100.00% uptime

Not used in Standard routing:Why these endpoints are not used

Input $0.375 / Output $1.87550% off

Flex

76 tok/s100.00% uptime

Input $0.375 / Output $1.87550% off

Flex

82 tok/s93.85% uptime

Input $0.825 / Output $4.12550% off

111 tok/s99.97% uptime

Input $1.35 / Output $6.7550% off

Priority

67 tok/s99.71% uptime

Input $1.35 / Output $6.7550% off

Priority

34 tok/s99.77% uptime

Pricing

The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.

Weighted average

Input

$0.4181/M tokens

Output

$4.65/M tokens

Gemini 3.8 Flash — Price history

Price history

No effective price data for this selection

$0.413$4.9968.0%28.7%

$0.445$4.9358.8%61.7%

Not used in Standard routing:Why these endpoints are not used

Flex

$0.198$1.8854.3%7.2%

Flex

$0.244$2.0141.4%1.2%

$0.437$4.1555.2%0.6%

Priority

$0.486$7.1373.4%0.2%

Priority

$1.667$12.5225.4%0.2%

ProviderInput / Output /M

Token share

Effective in $0.413 / Effective out $4.99

Token share

28.7%

Effective in $0.445 / Effective out $4.93

Token share

61.7%

Not used in Standard routing:Why these endpoints are not used

Effective in $0.198 / Effective out $1.88

FlexToken share

7.2%

Effective in $0.244 / Effective out $2.01

FlexToken share

1.2%

Effective in $0.437 / Effective out $4.15

Token share

0.6%

Effective in $0.486 / Effective out $7.13

PriorityToken share

0.2%

Effective in $1.667 / Effective out $12.52

PriorityToken share

0.2%

Performance

Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).

Throughput

111tok/s

P50, best across providers

Latency

1.83s

P50, best provider

Reasoning effortLatency / throughput

AutoExacto benchmarks

GPQA DiamondTAU-BenchGoogle Vertex94.9%78.0%

Uptime

Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.

Uptime (3d)The model was reachable. Request routed to a provider.

100.00%

Availability (3d)The model returned inference from any provider. Errors and empty responses count against it.

99.81%

Availability over the last 3 days

Last 72 hours

Availability 99.81%

3 Days Ago2 Days AgoYesterdayNow

Availability over the last 24 hours

OpenRouter Availability

99.83%

Without Routing

98.17%

When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and provider routing options.

Benchmarks

Scores on standardized evaluations and this model’s own outputs on shared benchmark prompts. Higher percentages are better — and each rank shows where this model lands among the models evaluated on that source.

Benchmark score summary for Google: Gemini 3.8 Flash (Artificial Analysis, Design Arena and Vals AI)

SourceBenchmarkScore

Artificial AnalysisGemini 3.8 Flash (Low) Intelligence Index22.5

Artificial AnalysisGemini 3.8 Flash (Low) Coding Index73.5

Artificial AnalysisGemini 3.8 Flash (Low) Agentic Index32.6

Artificial AnalysisGemini 3.8 Flash (Low) GPQA Diamond92.0%

Artificial AnalysisGemini 3.8 Flash (Low) HLE37.1%

Artificial AnalysisGemini 3.8 Flash (Low) AA-LCR80.7%

Artificial AnalysisGemini 3.8 Flash (Low) τ-Bench Banking33.2%

Artificial AnalysisGemini 3.8 Flash (Low) GDPval-AA40.7%

Artificial AnalysisGemini 3.8 Flash (Low) CritPt4.0%

Artificial AnalysisGemini 3.8 Flash (Low) SciCode55.0%

Artificial AnalysisGemini 3.8 Flash (Low) Terminal-Bench 4.010.1%

Artificial AnalysisGemini 3.8 Flash (Low) Terminal-Bench 2.183.1%

Artificial AnalysisGemini 3.8 Flash (Low) AA-Omniscience Accuracy48.3%

Artificial AnalysisGemini 3.8 Flash (Low) AA-Omniscience Non-Hallucination Rate31.9%

Artificial AnalysisGemini 3.8 Flash (High) Intelligence Index30.8

Artificial AnalysisGemini 3.8 Flash (High) Coding Index76.3

Artificial AnalysisGemini 3.8 Flash (High) Agentic Index40.2

Artificial AnalysisGemini 3.8 Flash (High) GPQA Diamond95.3%

Artificial AnalysisGemini 3.8 Flash (High) HLE47.8%

Artificial AnalysisGemini 3.8 Flash (High) AA-LCR81.3%

Artificial AnalysisGemini 3.8 Flash (High) τ-Bench Banking44.9%

Artificial AnalysisGemini 3.8 Flash (High) GDPval-AA47.1%

Artificial AnalysisGemini 3.8 Flash (High) CritPt18.3%

Artificial AnalysisGemini 3.8 Flash (High) SciCode56.6%

Artificial AnalysisGemini 3.8 Flash (High) Terminal-Bench 4.019.7%

Artificial AnalysisGemini 3.8 Flash (High) Terminal-Bench 2.187.6%

Artificial AnalysisGemini 3.8 Flash (High) AA-Omniscience Accuracy50.5%

Artificial AnalysisGemini 3.8 Flash (High) AA-Omniscience Non-Hallucination Rate44.9%

Artificial AnalysisGemini 3.8 Flash (Medium) Intelligence Index30.1

Artificial AnalysisGemini 3.8 Flash (Medium) Coding Index74.1

Artificial AnalysisGemini 3.8 Flash (Medium) Agentic Index39.7

Artificial AnalysisGemini 3.8 Flash (Medium) GPQA Diamond93.5%

Artificial AnalysisGemini 3.8 Flash (Medium) HLE42.1%

Artificial AnalysisGemini 3.8 Flash (Medium) AA-LCR84.0%

Artificial AnalysisGemini 3.8 Flash (Medium) τ-Bench Banking45.8%

Artificial AnalysisGemini 3.8 Flash (Medium) GDPval-AA46.6%

Artificial AnalysisGemini 3.8 Flash (Medium) CritPt12.3%

Artificial AnalysisGemini 3.8 Flash (Medium) SciCode55.1%

Artificial AnalysisGemini 3.8 Flash (Medium) Terminal-Bench 4.019.7%

Artificial AnalysisGemini 3.8 Flash (Medium) Terminal-Bench 2.183.9%

Artificial AnalysisGemini 3.8 Flash (Medium) AA-Omniscience Accuracy49.6%

Artificial AnalysisGemini 3.8 Flash (Medium) AA-Omniscience Non-Hallucination Rate48.5%

Design ArenaGemini 3.8 Flash Agents Arena Androidnative Elo1088

Design ArenaGemini 3.8 Flash Agents Arena Blender3d Elo1183

Design ArenaGemini 3.8 Flash Agents Arena Codetovideo Elo1194

Design ArenaGemini 3.8 Flash Agents Arena Full Stack Elo1216

Design ArenaGemini 3.8 Flash Agents Arena Godotgamedev Elo1247

Design ArenaGemini 3.8 Flash Agents Arena Mobile Apps Elo1225

Design ArenaGemini 3.8 Flash Agents Arena Pdf Elo1204

Design ArenaGemini 3.8 Flash Agents Arena Python-Pptxslides Elo1149

Design ArenaGemini 3.8 Flash Agents Arena Spreadsheet Elo1031

Design ArenaGemini 3.8 Flash Agents Arena Webapps Elo1214

Design ArenaGemini 3.8 Flash Models Arena 3D Elo1275

Design ArenaGemini 3.8 Flash Models Arena Asciiart Elo1261

Design ArenaGemini 3.8 Flash Models Arena Code Categories Elo1292

Design ArenaGemini 3.8 Flash Models Arena Data Visualization Elo1274

Design ArenaGemini 3.8 Flash Models Arena Game Development Elo1296

Design ArenaGemini 3.8 Flash Models Arena SVG Elo1289

Design ArenaGemini 3.8 Flash Models Arena UI Component Elo1303

Design ArenaGemini 3.8 Flash Models Arena Website Elo1291

Vals AIGemini 3.8 Flash Access to Justice Benchmark67.1%

Vals AIGemini 3.8 Flash BioMysteryBench62.2%

Vals AIGemini 3.8 Flash Code Migration v1.138.6%

Vals AIGemini 3.8 Flash CyberBench v1.143.8%

Vals AIGemini 3.8 Flash EMB v1.172.9%

Vals AIGemini 3.8 Flash Finance Agent v2.165.8%

Vals AIGemini 3.8 Flash Vals Finance Index60.5%

Vals AIGemini 3.8 Flash Financial Coaching Bench66.3%

Vals AIGemini 3.8 Flash GPQA Diamond94.4%

Vals AIGemini 3.8 Flash HLAB-Vals-1.112.1%

Vals AIGemini 3.8 Flash Immigration63.4%

Vals AIGemini 3.8 Flash IOI56.9%

Vals AIGemini 3.8 Flash LiveCodeBench89.5%

Vals AIGemini 3.8 Flash LegalBench87.0%

Vals AIGemini 3.8 Flash Vals Legal Index69.8%

Vals AIGemini 3.8 Flash Legal Research Bench38.9%

Vals AIGemini 3.8 Flash MedCode48.1%

Vals AIGemini 3.8 Flash MedScribe84.5%

Vals AIGemini 3.8 Flash MMLU Pro90.2%

Vals AIGemini 3.8 Flash MMMU Pro89.1%

Vals AIGemini 3.8 Flash MortgageTax65.3%

Vals AIGemini 3.8 Flash MysteryMechanism36.5%

Vals AIGemini 3.8 Flash ProgramBench1.0%

Vals AIGemini 3.8 Flash ProofBench v1.148.0%

Vals AIGemini 3.8 Flash Public Benefits Bench v1.165.3%

Vals AIGemini 3.8 Flash Vals RSI Index19.9%

Vals AIGemini 3.8 Flash SAGE88.2%

Vals AIGemini 3.8 Flash Vals Science Index35.8%

Vals AIGemini 3.8 Flash SkillsBench58.0%

Vals AIGemini 3.8 Flash SRE Bench11.1%

Vals AIGemini 3.8 Flash SWE-bench80.0%

Vals AIGemini 3.8 Flash SWE-bench Multimodal40.8%

Vals AIGemini 3.8 Flash Tax Agent Bench32.1%

Vals AIGemini 3.8 Flash TaxEval v274.4%

Vals AIGemini 3.8 Flash Terminal-Bench 4.019.7%

Vals AIGemini 3.8 Flash Terminal-Bench-Science8.6%

Vals AIGemini 3.8 Flash Vals Index56.0%

Vals AIGemini 3.8 Flash Vibe Code Bench 1-10018.8%

Vals AIGemini 3.8 Flash Vibe Code Bench v1.178.7%

Apps

Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.

1.

Hermes Agent

Hermes Agent is an open-source, self-improving AI agent by Nous Research that runs persistently with memory across sessions, and builds reusable skills from experience. It comes with 40+ built-in tools, including web search, browser automation, and vision, plus scheduled automations and subagents.

392Btokens

2.

HighLevel

new

358Btokens

3.

ISEKAI ZERO

AI adventures. Travel with your favorite characters

139Btokens

4.

pi

There are many coding agents, but this one is yours.

109Btokens

5.

OpenClaw

OpenClaw is an open-source AI agent that connects to your messaging apps and takes real actions on your behalf, from running commands and browsing the web to managing files and sending emails.

76.5Btokens

Rest of OpenRouter

68.2B

Top 5 Apps

11.7B

Activity

Token volume and request traffic to this model over time.

API

Drop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.

Google: Gemini Flash versions

Not currently available through EU in-region routing. In-region routing docs

Documentation

Guides, cookbooks, blog posts, and tools for building with Google: Gemini 3.8 Flash on OpenRouter.

Reranking in RAG: How to Fix Bad Retrieval with a RerankerBlog post

How to Benchmark LLMsBlog post

How to Stop an AI Agent Stuck in an Infinite LoopBlog post

MCP vs CLI vs Skills: What Each Costs in ContextBlog post

How to Get Structured Output from Any LLM: JSON Schema, Tool Calls, or a Decision ModelBlog post

Best OCR Model for Scanned Documents, MeasuredBlog post

How to Extract Tables from a PDF with an LLMBlog post

Frequently asked questions

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.

The Gemini 3.8 Flash documentation on OpenRouter includes the Reranking in RAG: How to Fix Bad Retrieval with a Reranker blog post, the How to Benchmark LLMs blog post, the How to Stop an AI Agent Stuck in an Infinite Loop blog post, the MCP vs CLI vs Skills: What Each Costs in Context blog post, the How to Get Structured Output from Any LLM: JSON Schema, Tool Calls, or a Decision Model blog post, the Best OCR Model for Scanned Documents, Measured blog post, and the How to Extract Tables from a PDF with an LLM blog post. The quickstart covers API keys and your first request, and the API reference lists every parameter.

Gemini 3.8 Flash costs $0.75/M input tokens and $3.75/M output tokens, with separate rates for Cache Read at $0.075/M tokens, Cache Write at $0.04167/M tokens, Image Input at $0.75/M tokens, Input Audio at $0.75/M tokens, Input Audio Cache at $0.075/M tokens, and Web Search at $14.00/1K calls.

Gemini 3.8 Flash has a 1,048,576 token context window. It supports up to 65,536 completion tokens.

Yes. Gemini 3.8 Flash accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.

Gemini 3.8 Flash accepts text, images, video, files such as PDFs, and audio as input and returns text.

Gemini 3.8 Flash is served by 2 providers on OpenRouter: Google AI Studio and Google Vertex. Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.

Gemini 3.8 Flash was released on September 2, 2026.

正文以来源语言显示。

参考来源

← 返回每日简报