Gemini 3.7 Flash API

NewLLM

Gemini 3.7 Flash is the previous Flash snapshot, kept at the same 1M context, capabilities, and rate as 3.8 Flash.

Input · USD / 1M tokens$0.45 USDOfficial price~~$0.75 USD~~
Output · USD / 1M tokens$2.25 USDOfficial price~~$3.75 USD~~
Savings40%
Context1M
View tool pricing and billing rules

What are we making today?

Bring an idea, a question, or a first draft.

What this model is for

Gemini 3.7 Flash is the previous Flash snapshot, kept at the same 1M context, capabilities, and rate as 3.8 Flash.

Gemini 3.7 Flash is the previous Flash snapshot in Google's Gemini 3.x line, kept in the OmniaKey catalog with the same 1M-token context window, image input, tool use and coding clients as its successor. Its id is gemini-3.7-flash. It carries the catalog's fast label and the new-model flag, and it lists the identical rate triple as Gemini 3.8 Flash, which is the mainline entry today. Google's 3.x Flash tiers sit below the Pro models and above the Lite ids.

Because the rate and the capability list match the successor, this id is a pinning decision rather than a cost decision. Nothing in the catalog makes a 3.7 call cheaper than a 3.8 call, so a swap buys a newer snapshot and nothing else. Keep 3.7 where an evaluation script, a stored result set or a vendored client configuration was built against it and the replay work has not been paid for yet. Move to 3.8 when a fresh baseline is acceptable, and go to Gemini 4 Argon only when a task needs a Pro model or one request has to carry a whole document set.

Treat the published details carefully. The thinking levels, the exact input and output ceilings and the caching minimum appear on Google's 3.8 material, and the catalog entry for 3.7 lists capabilities and a rate without repeating them. Before inheriting any of those settings for this id, confirm them against Google's own pages and then measure the answer on your own prompts.

Keep a deployment on it when it already pins the 3.7 id; new work should move to 3.8 Flash at the same price.

Best-fit coding workloads

Snapshot A/B on stored prompts

Take a frozen set of 50 recorded prompts with known good answers and run it against gemini-3.7-flash and gemini-3.8-flash, scoring both with the same script. Since the catalog rate matches, any difference you see comes from the model revision. Practical tip: store prompts, parameters and scores together so a later run can repeat the comparison.

Nightly ticket classification

Classify a queue of support tickets every night with one fixed prompt and one fixed id, and keep weekly counts per category. A pinned snapshot keeps the series comparable, which matters more than a newer revision when the output feeds a dashboard. Practical tip: sample a hundred rows for human review each week and track the disagreement rate.

Pinned client maintenance

Keep a coding assistant built on Cursor, Cline or Aider pointed at the id it already uses while a wider upgrade is planned. Changing the model string in an editor plugin is a configuration edit, not a code change. Practical tip: record which id each developer profile uses, so a support question can be answered without guesswork.

Hands-on notes

Checked on 2026-10-27 through the OmniaKey API with one standard coding prompt (median of 2 runs) and three capability probes.

Output tokens for our standard coding prompt, reasoning included: 1,958
Cost of that output at current prices: $0.0044
JSON mode: Passed
Tool calling: Passed
Image input: Passed

Technical specification

Catalog facts and integration details verified for this route.

Model ID: gemini-3.7-flash
Context: 1M
Inputs and tools: Vision · Tool use · Fast responses
API endpoint: /v1/chat/completions
Cache hit / 1M: $0.045 USD
Coding clients: Cursor · Cline · Aider

Limits to plan around

The catalog gives this id the same input, output and cached-input rates as Gemini 3.8 Flash, so moving to the newer snapshot neither raises nor lowers the listed charge. There is no billing argument for staying, only a reproducibility one.
The operational details published for 3.8, including thinking levels and the exact token ceilings, are not repeated on this entry. Confirm each setting on Google's pages before assuming the 3.8 defaults apply here.
A snapshot still shares the family's 1M listing, and that is a ceiling rather than a working size. A prompt that large needs retrieval and trimming to be useful, and your client may reject it long before the model does.
Two ids in one evaluation break comparability. If half a week's runs use 3.7 and half use 3.8, the score change mixes a model revision with a change of test conditions.

Compare nearby models

Compare workload fit and current listed token rates before choosing a route.

Model	Best for	Context	Input	Output
Gemini 3.7 FlashThis page gemini-3.7-flash	Reproducing earlier Flash results, or keeping a client pinned to the 3.7 id without replay work.	1M	$0.45 USD Official price $0.75 USD	$2.25 USD Official price $3.75 USD
Gemini 3.8 Flash gemini-3.8-flash	A fresh baseline on the current mainline Flash snapshot, at the same catalog rates.	1M	$0.45 USD Official price $0.75 USD	$2.25 USD Official price $3.75 USD
Gemini 4 Argon gemini-4-argon	Pro-tier work, one request per document set, or video and audio input.	1M	$0.8 USD Official price $2 USD	$4 USD Official price $10 USD

SDK quickstarts

Use one base URL and model ID from cURL, Python, or TypeScript.

Streaming chat completion

curl https://api.omniakey.com/v1/chat/completions \
  -H "Authorization: Bearer $OMNIAKEY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-3.7-flash",
    "messages": [
      {"role": "user", "content": "Review this pull request"}
    ],
    "stream": true
  }'

Estimate token cost

Enter expected input and output tokens to compare current listed rates.

Input tokensOutput tokens

OmniaKey estimate: $0.09
Official-rate estimate: $0.15
Estimated savings: $0.06

Estimate covers input and output tokens only. Cache usage, retries, tools, and client-side token accounting can change the final charge.

Related coding guides

Deeper selection, setup, and troubleshooting guidance for this model family.

Gemini 3.8 Flash API Pricing and Agent Cost

Google's introductory rate is $0.75 input and $3.75 output per million tokens, but retries and thinking decide the cost of an accepted result.

Read guide

Gemini 4 Release Date: Confirmed, Not Released

The Gemini 4 release date came in stages: Google announced Gemini 4 Argon on September 30, 2026 and opened it to paid API developers on October 13, 2026, then announced Gemini 4 Carbon on October 28, 2026 with no developer date yet.

Read guide

Sources and verification

Model facts and links were checked on 2026-10-27. Provider documentation remains the authority for upstream behavior.

Frequently asked questions

How much does Gemini 3.7 Flash cost on OmniaKey?

OmniaKey charges $0.45 USD per 1M input tokens and $2.25 USD per 1M output tokens, in USD. Official prices are $0.75 USD and $3.75 USD; you save 40%.

What is the context window for Gemini 3.7 Flash?

The listed context window is 1M. Actual usable limits can also depend on the client and upstream request format.

What is Gemini 3.7 Flash?

It is the previous Flash snapshot in Google's Gemini 3.x line, listed on OmniaKey with a 1M-token context window, image input, tool use, the fast label and three coding clients. Its catalog rates match Gemini 3.8 Flash exactly, so the two differ by snapshot rather than by price tier.

Should I call 3.7 Flash or 3.8 Flash?

For new work, use the current mainline id, gemini-3.8-flash. Keep gemini-3.7-flash when an evaluation set, a stored result series or a pinned client was built against it and you have not replayed the tests. The listed rate is the same, so the choice is about reproducing earlier behaviour.

How do I call Gemini 3.7 Flash?

POST to https://api.omniakey.com/v1/chat/completions with gemini-3.7-flash in the model field. The request body is OpenAI-compatible, so text, image parts and tool definitions go in the messages array. One OmniaKey key covers it, and the id is case-sensitive, so copy the string instead of typing it from memory.

Does it have the same limits as Gemini 3.8 Flash?

The catalog lists the same 1M context and the same capability set. Google's model page for 3.8 gives 1,048,576 input tokens and 65,536 output tokens; those figures are published for 3.8, not for this entry. Check Google's pages for the 3.7 limits and test one long request before you budget around them.

Why keep an older Flash snapshot at all?

Reproducibility. A stored evaluation set, a dashboard series or a support answer that quotes past results stays meaningful only if the model id behind those runs is still callable. Once a fresh baseline is acceptable, the pinned id can be retired in one configuration change.