Gemini 3.7 Flash API: Quickstart, Pricing Calculator & Migration Checklist

Direct answer: Gemini 3.7 Flash is generally available under the stable model ID gemini-3.7-flash. Google positions it for coding and agent workflows, with a 1,048,576-token input limit, up to 65,536 output tokens, and low, medium or high thinking levels.[1][2]

Use the copy-and-paste examples below to make a first request, estimate token cost, or migrate an existing Gemini 3.x app.

Pricing alert: Google lists an introductory rate of $0.75 per 1 million input tokens and $3.75 per 1 million output tokens through December 31, 2026. The listed rates become $1.50 input and $7.50 output per million tokens on January 1, 2027.[1][3]

Gemini 3.7 Flash at a glance

Item	Gemini 3.7 Flash
Status	Generally available (production-ready)
Model ID	gemini-3.7-flash
Inputs	Text, image, video, audio and PDF
Output	Text
Input limit	1,048,576 tokens
Maximum output	65,536 tokens
Thinking levels	low, medium (default), high
Intro API price	$0.75 input / $3.75 output per 1M tokens
Standard price from Jan. 1, 2027	$1.50 input / $7.50 output per 1M tokens

These model specifications and prices come from Google’s launch announcement and current Gemini API documentation.[1][2][3]

Quickstart: call Gemini 3.7 Flash

1. Get an API key

Create a key in Google AI Studio, then store it as an environment variable instead of placing the key directly in source code.

export GEMINI_API_KEY="your-api-key"

2. Python example

Install Google’s Gen AI SDK:

pip install -U google-genai

Then make a request through the Interactions API, which Google currently recommends for access to its latest models and features.[2]

from google import genai

client = genai.Client()

interaction = client.interactions.create(
    model="gemini-3.7-flash",
    input="Create a five-step migration plan for this Python service."
)

print(interaction.output_text)

3. JavaScript example

npm install @google/genai

import { GoogleGenAI } from "@google/genai";

const client = new GoogleGenAI({});

const interaction = await client.interactions.create({
  model: "gemini-3.7-flash",
  input: "Review this function and suggest safer error handling.",
});

console.log(interaction.output_text);

4. cURL example

curl "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H "Content-Type: application/json" \
  -X POST \
  -d '{
    "model": "gemini-3.7-flash",
    "input": "Return a concise launch checklist for a small website."
  }'

The request shape and stable model ID above follow Google’s official Gemini 3.7 Flash quickstart.[2]

Gemini 3.7 Flash cost calculator

Use this formula for standard text requests:

Estimated cost = (input tokens / 1,000,000 × input rate)
               + (output tokens / 1,000,000 × output rate)

Worked examples

Usage	Intro price through Dec. 31, 2026	Price from Jan. 1, 2027
100,000 input + 20,000 output tokens	$0.15	$0.30
1M input + 250,000 output tokens	$1.6875	$3.375
10M input + 2M output tokens	$15.00	$30.00

These are arithmetic estimates using Google’s listed token rates; actual billing can also include tool use, retrieved content, agent loops or other billable usage.[3]

What changed from Gemini 3.6 Flash?

Google reports gains in software engineering, web development, document reasoning and business workflow benchmarks. Its launch examples include FrontierCode 1.1 Main at 43.6% versus 34.4% for 3.6 Flash, DeepSWE v1.1 at 65.3% versus 49.0%, and AutomationBench at 30.4% versus 17.0%.[1]

Treat benchmark results as directional rather than a guarantee for your workload. Run the same prompts and evaluation set against both model versions before changing production traffic.

Migration checklist for existing Gemini apps

Google’s current migration guidance calls for more than changing the model name.[2]

• [ ] Change the model string to gemini-3.7-flash.
• [ ] Remove temperature, top_p and top_k from generation configuration.
• [ ] Replace thinking_budget with thinking_level.
• [ ] Remove candidate_count, which is unsupported in Gemini 3.x.
• [ ] Use server-side previous_interaction_id for multi-turn conversations.
• [ ] Remove prefilled model turns.
• [ ] Re-test function calls and malformed-call error handling.
• [ ] Preserve thought signatures where your integration requires them.
• [ ] Compare quality, latency and total token use on your own evaluation set.
• [ ] Keep the model ID in an environment variable so rollback is a configuration change.

Example environment setup

export GEMINI_MODEL="gemini-3.7-flash"

import os
from google import genai

client = genai.Client()
model = os.environ.get("GEMINI_MODEL", "gemini-3.7-flash")

interaction = client.interactions.create(
    model=model,
    input="Summarize the deployment risks in this change request."
)
print(interaction.output_text)

Which thinking level should you use?

Gemini 3.7 Flash supports low, medium and high, with medium as the documented default.[2]

• Low: start here for extraction, classification and short transformations where speed matters.
• Medium: use as the general baseline for coding, planning and multi-step work.
• High: test for difficult debugging or planning tasks, then compare the quality improvement with latency and token use.

Do not assume that a higher setting is automatically better for every request. Measure success rate, response time, input/output tokens and retry frequency on representative tasks.

Five useful test prompts

1. Code review: “Review this pull request for correctness, security risks and missing tests. Return findings by severity, then provide minimal patches.”
2. UI implementation: “Recreate this supplied design as a responsive component. List assumptions before writing code and include accessibility checks.”
3. Document extraction: “Extract obligations, dates, owners and unresolved risks from these PDFs. Return valid JSON and cite the source filename for every item.”
4. Agent planning: “Create a step-by-step execution plan, identify which steps require tools, and stop for confirmation before irreversible actions.”
5. Migration test: “Compare these two outputs against the rubric. Flag regressions, unsupported claims and formatting failures.”

FAQ

Is Gemini 3.7 Flash generally available?

Yes. Google documents gemini-3.7-flash as generally available and ready for production use.[2]

What is the Gemini 3.7 Flash model ID?

The stable model ID is gemini-3.7-flash.[2]

How much does Gemini 3.7 Flash cost?

Through December 31, 2026, Google lists $0.75 per million input tokens and $3.75 per million output tokens. From January 1, 2027, the listed rates are $1.50 and $7.50 respectively.[1][3]

Does Gemini 3.7 Flash accept PDFs and images?

Yes. Google’s model documentation lists text, image, video, audio and PDF as supported inputs; output is text.[2]

Should I remove temperature and top-p when migrating?

Google’s Gemini 3.7 Flash migration checklist says to remove temperature, top_p and top_k, replace thinking_budget with thinking_level, and remove candidate_count.[2]

Is Gemini 3.7 Flash better than 3.6 Flash?

Google reports higher results on several coding, web-development and workflow benchmarks.[1] The practical answer depends on your prompts, tools and evaluation criteria, so run a controlled comparison before switching all production traffic.

Bottom line

Gemini 3.7 Flash is a strong candidate for teams that need one fast model for coding, document work and agentic execution. Start with the official model ID, use the recommended Interactions API, calculate costs against real token usage, and complete the migration checklist before sending production traffic.[1][2][3]

Sources

[1] https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-3-7-flash — Introducing Gemini 3.7 Flash
[2] https://ai.google.dev/gemini-api/docs/latest-model — What’s new in Gemini 3.7 Flash — API guide
[3] https://ai.google.dev/gemini-api/docs/pricing — Gemini Developer API pricing