Gemini API
UnificAlly serves 15 Google models through one API key: 4 language models, 4 image models, 2 video models and 5 text to speech models. Language models start at $0.225 per 1M input tokens (Gemini 3.8 Flash). Image, video and text to speech models start at $0.000105/sec (Gemini 3.8 Flash Lite TTS) and $0.02 per image (Nano Banana 2.1). Everything is billed from a prepaid balance with no subscription.
Call the language models in the OpenAI format at /v1/chat/completions or the Anthropic format at /v1/messages, with streaming and tool use; the image, video and text to speech models take POST /v1/tasks with a model ID and an input object.
UnificAlly is an independent API, not affiliated with or endorsed by Google: you use a UnificAlly API key and balance, not an account with Google.
All 15 Google models on UnificAlly
| Model | Type | Starting rate |
|---|---|---|
| Gemini 3.8 Flash API LLM 1M context, four reasoning efforts, 73.7% DeepSWE v1.1, 89.4% Terminal-Bench 2.1, 305 tok/s. | LLM | $0.225 in / $1.125 out per 1M tokens |
| Gemini 3.7 Flash API LLM 1M context, 64k output, text/image/audio/video input, ~340 tok/s, 97% MRCR long-context recall. | LLM | $0.225 in / $1.125 out per 1M tokens |
| Gemini 3.6 Flash API LLM 1M context, 65k output, multimodal input, 83% OSWorld-Verified, 234 tok/s output speed. | LLM | $0.45 in / $2.25 out per 1M tokens |
| Gemini 3.5 Flash API LLM 1M context, 5 reasoning efforts incl. minimal, 83.6% MCP Atlas, 57.9% Finance Agent v2, 189 tok/s. | LLM | $0.45 in / $2.7 out per 1M tokens |
| Nano Banana 2.1 API text-to-image Nano Banana 2.1: 1K, 2K, or 4K output, 5 aspect ratios plus auto, up to 10 references. | text-to-image | from $0.02 per image |
| Nano Banana 2 API text-to-image Gemini image gen with 1K, 2K, or 4K resolution control and 10 aspect ratios. | text-to-image | from $0.034 per image |
| Nano Banana 2 Lite API text-to-image Sub-2s generation, 14 aspect ratios, 1K output, up to 14 reference images. | text-to-image | $0.02 per image |
| Nano Banana Pro API text-to-image Gemini 3 image gen with 1K, 2K, or 4K, 10 aspect ratios, and reference image support. | text-to-image | from $0.07 per image |
| Veo 3.1 API text-to-video 4 model variants, 720p–4K, lip-synced dialogue and SFX in one call. | text-to-video | from $0.075 per video |
| Gemini Omni Flash 1.1 API text-to-video Text, image, and character references with voice presets; 4-10s videos plus conversational editing. | text-to-video | from $0.11 per video |
| Gemini 3.8 Flash Lite TTS API text-to-speech The high-volume Gemini 3.8 voice, Elo 1241 on Artificial Analysis TTS, 100+ languages. | text-to-speech | $0.000105/sec |
| Gemini 3.8 Flash TTS API text-to-speech Elo 1267 on Artificial Analysis TTS, vocal bursts, two-speaker dialogue, 130+ languages. | text-to-speech | $0.000158/sec |
| Gemini 3.1 Flash TTS API text-to-speech Inline audio tags, two-speaker dialogue, 30 voices, 70+ languages. | text-to-speech | $0.00035/sec |
| Gemini 2.5 Pro TTS API text-to-speech High-quality narration with 30 voices, dialogue mode, and 70+ languages. | text-to-speech | $0.00035/sec |
| Gemini 2.5 Flash TTS API text-to-speech Fastest, lowest cost per second, with 30 voices and dialogue mode. | text-to-speech | $0.000175/sec |
How to call the Gemini API
Use your UnificAlly API key and any model ID from the model pages. Existing OpenAI clients only need the base URL changed to https://api.unifically.com/v1.
OpenAI format: POST /v1/chat/completions
curl -X POST https://api.unifically.com/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "google/gemini-3.8-flash",
"messages": [
{"role": "user", "content": "Summarize this document in three bullet points."}
]
}'Image and video models: POST /v1/tasks
# 1. Start a task
curl -X POST https://api.unifically.com/v1/tasks \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "google/nano-banana-2",
"input": {
"prompt": "A photorealistic portrait in golden hour lighting",
"aspect_ratio": "3:4"
}
}'
# 2. Read the result with the task_id from step 1
curl https://api.unifically.com/v1/tasks/TASK_ID \
-H "Authorization: Bearer YOUR_API_KEY"Gemini API FAQ
- Is this the official Gemini API from Google?
- No. UnificAlly is an independent API that serves Google's models; it is not affiliated with or endorsed by Google. You sign up with UnificAlly, create a UnificAlly API key and pay from a UnificAlly balance. The request formats are the same OpenAI and Anthropic formats, so existing code needs a new base URL, API key and model ID.
- How much does the Gemini API cost on UnificAlly?
- Each model has its own rate, shown in the table above and on the pricing page. The lowest language model rate right now is $0.225 per 1M input tokens (Gemini 3.8 Flash). Image, video and text to speech models start at $0.000105/sec (Gemini 3.8 Flash Lite TTS) and $0.02 per image (Nano Banana 2.1). You pay per token or per generation from a prepaid balance: no subscription, no monthly minimum, and the balance does not expire. Failed generations are not charged.
- Are Nano Banana and Veo part of the Gemini API here?
- Yes. Nano Banana is the name of Google's Gemini image models, and Google's own Gemini API also serves Veo, so this page lists every Google model on UnificAlly. Nano Banana 2.1, Nano Banana 2, Nano Banana 2 Lite, Nano Banana Pro, Veo 3.1 and Gemini Omni Flash 1.1 run through POST /v1/tasks, as do the 5 Gemini text to speech models; the Gemini language models use /v1/chat/completions or /v1/messages.
- Can I call the Gemini models with the OpenAI SDK?
- Yes. Point your OpenAI client at https://api.unifically.com/v1, use a UnificAlly API key and set model to an ID such as google/gemini-3.8-flash (each model page lists its ID). Anthropic SDK code works too on /v1/messages. The API speaks the OpenAI and Anthropic formats, not Google's generateContent format, so code written for Google's own SDK moves to one of those clients.
- Can I try the Gemini models before topping up?
- Yes. A new UnificAlly account starts with a $0.50 signup balance, and every model page has a browser playground that runs on it, so you can compare models before writing any code.