gemini-3.7-flash
Gemini 3.7 Flash is the next iteration in the Gemini 3 series of highly-capable, natively multimodal, reasoning models.
Modalities
In / Out Price
40% off$0.45/ 1M
$0.75 Official price
$2.25/ 1M
$3.75 Official price
Context
1.048576M
Released
Aug 13, 2026
API call examples
POST /v1/chat/completionscurl --request POST 'https://api.tokatlas.com/v1/chat/completions' \
--header "Authorization: Bearer $UNIROUTER_API_KEY" \
--header 'Content-Type: application/json' \
--data '{
"model": "gemini-3.7-flash",
"messages": [
{
"role": "user",
"content": "Explain quantum entanglement."
}
],
"temperature": 0.7
}'Providers & routes
Current provider · Google
Platform configured price · Group: default
USD / 1M tokens. Each row shows input and output from the same tier. Displayed rates use the lowest available group multiplier; your API key group may differ.
Gemini 3.7 Flash for verifiable media analysis
Use it to organize long documents and visual materials into evidence-backed findings. Set a consistent output format before comparing sources.
Multimodal context
Provide documents, visuals, or media with timestamps, page numbers, or source labels.
Development and UI
Compare screenshots with requirements and produce testable acceptance criteria.
Structured fields
Include source locations and unresolved points in summaries.
Three thinking levels
Compare low, medium, and high thinking on the same evaluation set.
About
Applications and context
3.7 Flash accepts text, images, audio or video, and PDFs, returning text. Segment long videos and documents first, then connect findings through consistent time or page indexes.
Comparison with 3.6 Flash
Keep media, prompts, and output formats fixed when comparing extraction, video localization, and coding suggestions. For tool workflows, validate function selection and returned structures in the application.
Thinking levels and route coverage
Listed thinking levels are low, medium, and high with medium as default. Verify generation or live-conversation support separately. Test batch serving and caching when the integration depends on them.
Separate interpretation from reporting. Identify screen regions or timestamps, then preserve evidence in JSON or checklists. When comparing with an older version, independently test compatibility with the tools and media fields the integration will use.
gemini-3.7-flash FAQ
Can Gemini 3.7 Flash summarize a PDF and video together?
It can understand multiple media types. After confirming that the route accepts mixed inputs, test a short clip and a few pages and ask the model to identify the relevant sections or timestamps.
How should I evaluate long-document performance before migrating?
Use documents with realistic lengths and layouts, include questions that span sections, and measure answer accuracy and response time.
Can I reuse native Gemini SDK examples as-is?
Start by checking the request format listed here. If you move to a compatible endpoint, test a minimal message first, then validate streaming and attachment fields.
What should change in my tests when moving from Gemini 3.6?
Keep the existing context tests, then rerun cases that require reading across sections, editing to an interface spec or completing several dependent steps.