Google: Gemini 3.8 Flash
google/gemini-3.8-flash
Our take
AI-assisted analysis based on the sources below. Suitability is an editorial assessment; test it on your own workload.
OpenRouter’s catalog describes google/gemini-3.8-flash as a Google multimodal model accepting text, image, video, file, and audio inputs and returning text. The listing reports a very large context window, mandatory reasoning enabled by default with selectable effort, and support for tools and structured outputs. Its catalog description claims gains over Gemini 3.7 Flash in software engineering, agentic work, and multi-step reasoning, while benchmark fields report measured results across design-arena tasks and coding, agentic, and intelligence indices.[1]
Strengths
- The catalog lists multimodal inputs of text, image, video, files, and audio, with text output.[1]
- The catalog reports a large context window plus controls for reasoning effort, tool calls, structured output, response format, seed, sampling, and stop sequences.[1]
- Reasoning is mandatory and enabled by default, with low, medium, and high effort settings.[1]
Limitations
- The only supplied source is the OpenRouter catalog; an independent model card was not available in this source pack, so capability and performance claims remain catalog-reported.[1]
- Benchmark entries are supplied without evaluation methodology here, so results should not be generalized without workload-specific validation.[1]
- Output is limited to text even though the listed inputs include images, video, files, and audio.[1]
Where to use it
- Consider evaluating it for assistants that combine image, audio, video, or file inspection with text responses; the modality support is catalog-reported and was not independently verified here.[1]
- It may be worth testing on long-document or multi-file analysis because the catalog reports a large context window; suitability still depends on task-specific trials.[1]
- It could be evaluated in software or agentic prototypes that use tools or structured text; these are catalog-suggested evaluation areas, not proven outcomes.[1]
Sources
- [1] OpenRouter catalog · Google: Gemini 3.8 FlashSources checked: Nov 13, 2026
About this model
Description below is the original text supplied by OpenRouter, shown in the source language.
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.
Specifications
- Model ID
- google/gemini-3.8-flash
- Canonical ID
- google/gemini-3.8-flash-20260902
- Creator
- Added to OpenRouter
- 02.09.2026
- Context length
- 1,048,576
- Modality
- text+image+file+audio+video->text
- Input modalities
- text, image, video, file, audio
- Output modalities
- text
- Token pricing
- USD per 1M tokens: in $0.825 / out $4.13
- Knowledge cutoff
- —
- Expiration date
- —
Tasks and capabilities
A grade appears only where the source supplies a matching measurement for that task, the model is active and the cohort holds at least 10 comparable models. The capability column lists facts the source declares — supported formats, reasoning and tool parameters — which are not measured quality.
| Task | Grade | Source-declared capability |
|---|---|---|
| General quality | BB — next 30% of the cohort | reasoning: mandatory · default medium · efforts: high, medium, low · parameters: include_reasoning, reasoning, reasoning_effort |
| Coding | SS — top 10% of the cohort | — |
| Agents | AA — next 20% of the cohort | parameters: structured_outputs, tool_choice, tools |
| Writing | No measurement | output: text |
| Reasoning | No measurement | reasoning: mandatory · default medium · efforts: high, medium, low · parameters: include_reasoning, reasoning, reasoning_effort |
| Document analysis | No measurement | input: file |
| Vision | No measurement | input: image |
| Image generation | No measurement | — |
| Audio | No measurement | input: audio |
| Video | No measurement | input: video |
| Embeddings | No measurement | — |
Evaluations
| Source | Benchmark | Value | Grade | Data retrieved |
|---|---|---|---|---|
| Artificial Analysis | Intelligence index | 30.8 | BB — next 30% of the cohort | 13.11.2026 |
| Artificial Analysis | Coding index | 76.3 | SS — top 10% of the cohort | 13.11.2026 |
| Artificial Analysis | Agentic index | 40.2 | AA — next 20% of the cohort | 13.11.2026 |
| Design Arena | agents · androidnativewin rate 34.6% · rank 35 | 1,086Elo | CC — below the B band | 13.11.2026 |
| Design Arena | agents · blender3dwin rate 48% · rank 6 | 1,192Elo | AA — next 20% of the cohort | 13.11.2026 |
| Design Arena | agents · codetovideowin rate 54.4% · rank 10 | 1,189Elo | BB — next 30% of the cohort | 13.11.2026 |
| Design Arena | agents · fullstackwin rate 47.7% · rank 13 | 1,220Elo | AA — next 20% of the cohort | 13.11.2026 |
| Design Arena | agents · godotgamedevwin rate 52.5% · rank 5 | 1,251Elo | AA — next 20% of the cohort | 13.11.2026 |
| Design Arena | agents · mobileappswin rate 54.4% · rank 11 | 1,228Elo | AA — next 20% of the cohort | 13.11.2026 |
| Design Arena | agents · pdfwin rate 53.2% · rank 15 | 1,201Elo | BB — next 30% of the cohort | 13.11.2026 |
| Design Arena | agents · python-pptxslideswin rate 40.8% · rank 22 | 1,142Elo | CC — below the B band | 13.11.2026 |
| Design Arena | agents · spreadsheetwin rate 44% · rank 22 | 1,112Elo | CC — below the B band | 13.11.2026 |
| Design Arena | agents · webappswin rate 46.5% · rank 16 | 1,219Elo | BB — next 30% of the cohort | 13.11.2026 |
| Design Arena | models · 3dwin rate 49.6% · rank 23 | 1,278Elo | AA — next 20% of the cohort | 13.11.2026 |
| Design Arena | models · asciiartwin rate 52.6% · rank 9 | 1,273Elo | AA — next 20% of the cohort | 13.11.2026 |
| Design Arena | models · codecategorieswin rate 50% · rank 18 | 1,293Elo | AA — next 20% of the cohort | 13.11.2026 |
| Design Arena | models · datavizwin rate 52.5% · rank 20 | 1,278Elo | AA — next 20% of the cohort | 13.11.2026 |
| Design Arena | models · gamedevwin rate 50.8% · rank 19 | 1,297Elo | AA — next 20% of the cohort | 13.11.2026 |
| Design Arena | models · svgwin rate 54.8% · rank 11 | 1,291Elo | AA — next 20% of the cohort | 13.11.2026 |
| Design Arena | models · uicomponentwin rate 52.4% · rank 17 | 1,306Elo | AA — next 20% of the cohort | 13.11.2026 |
| Design Arena | models · websitewin rate 49.5% · rank 19 | 1,292Elo | AA — next 20% of the cohort | 13.11.2026 |
How grades work
Each letter grade compares a model only against other models scored by the same source and metric in the current snapshot. S means the top 10%, A the next 20%, B the next 30%, C the rest. Models without a published measurement are shown as not rated — a supported format alone never implies measured quality.
- S — S — top 10% of the cohort
- A — A — next 20% of the cohort
- B — B — next 30% of the cohort
- C — C — below the B band
Providers
Optional live provider statistics from OpenRouter. Missing values are unknown, not zero.
| Provider | Context length | Price in /1M | Price out /1M | Latency p50 (30m, s) | Throughput p50 (30m, tok/s) | Uptime (30m) |
|---|---|---|---|---|---|---|
| Google AI Studio | 1,048,576 | $0.375 | $1.88 | — | — | 100% |
| 1,048,576 | $0.375 | $1.88 | — | — | 0% | |
| Google AI Studio | 1,048,576 | $0.750 | $3.75 | — | — | 100% |
| 1,048,576 | $0.750 | $3.75 | — | — | 100% | |
| 1,048,576 | $0.825 | $4.13 | — | — | 100% | |
| Google AI Studio | 1,048,576 | $1.35 | $6.75 | — | — | 100% |
| 1,048,576 | $1.35 | $6.75 | — | — | 100% |
Provider data retrieved: 13.11.2026
Related articles
- Google introduces Gemini 4 Hydrogen, a fast model for everyday and agentic workflowsGoogle launches Gemini 4 Hydrogen, a speed-focused model for coding and agentic work, across the Gemini app, Search, AI Studio, the Gemini API, and enterprise surfaces, with flat pricing over a 1 million-token context.
- ExpertCrowd-SWE: Expert-Authored Benchmark Puts Seven Frontier Coding Agents 35 Points ApartA Toptal ExpertCrowd technical report introduces EC-SWE, 100 engineer-written software tasks graded like maintainer review; seven frontier models span 35 points, with Claude Opus 5.5 and GPT-6 Astra statistically tied at the top.