Gemini API
Use Google Gemini through SandBase: Gemini 4 Carbon and Argon for the hardest reasoning, and Gemini 3.x Flash and Flash-Lite for fast multimodal work.
Explore the model lineup
Preview the output style, compare capabilities, then open any model for full specs, pricing, and a live Playground.
Gemini 4 Argon
Gemini 4 Argon is Google's flagship reasoning model and the first model of the Gemini 4 generation, generally available since October 13, 2026. Google trained it to keep context and precision across l...
Gemini 3.8 Flash
Gemini 3.8 Flash is a Flash-tier reasoning model in Google's Gemini 3 series, released in September 2026 and described by Google as its most intelligent Flash model as of release. Google positions it ...
Gemini 3.7 Flash
Gemini 3.7 Flash is a natively multimodal reasoning model in Google's Gemini 3 Flash line, released in August 2026, three weeks after 3.6 Flash. Google positions it as an everyday driver for developer...
Gemini 3.5 Flash Lite
Gemini 3.5 Flash-Lite is the low-latency, high-throughput tier of Google's Gemini 3.5 family, released in July 2026 and described by Google as its fastest 3.5-class model. It is optimized for subagent...
Gemini 3.1 Pro Preview
Gemini 3.1 Pro Preview is the Pro-tier reasoning model of Google's Gemini 3 series, released in February 2026 to refine Gemini 3 Pro with better thinking, improved token efficiency and more grounded, ...
Gemini 4 Carbon
Gemini 4 Carbon is Google's most capable Gemini 4 model, released on November 10, 2026 and positioned above Gemini 4 Argon. Google built it for long-horizon agentic work such as complex software engin...
Model lineup
Context window and positioning from each provider.
| Model | Context | Positioning |
|---|---|---|
| Gemini 3.8 Flash | 1049K | Gemini 3.8 Flash is a Flash-tier reasoning model in Google's Gemini 3 series, released in September 2026 and described by Google as its most… |
| Gemini 3.7 Flash | 1049K | Gemini 3.7 Flash is a natively multimodal reasoning model in Google's Gemini 3 Flash line, released in August 2026, three weeks after 3.6 Flash |
| Gemini 3.5 Flash Lite | 1049K | Gemini 3.5 Flash-Lite is the low-latency, high-throughput tier of Google's Gemini 3.5 family, released in July 2026 and described by Google as its… |
| Gemini 3.1 Pro Preview | 1049K | Gemini 3.1 Pro Preview is the Pro-tier reasoning model of Google's Gemini 3 series, released in February 2026 to refine Gemini 3 Pro with better… |
| Gemini 4 Argon | 1049K | Gemini 4 Argon is Google's flagship reasoning model and the first model of the Gemini 4 generation, generally available since October 13, 2026 |
| Gemini 4 Carbon | 1049K | Gemini 4 Carbon is Google's most capable Gemini 4 model, released on November 10, 2026 and positioned above Gemini 4 Argon |
Current Google Gemini models
SandBase serves Gemini on the OpenAI-compatible POST /v1/chat/completions. Gemini models are natively multimodal and accept audio input in addition to text and images.
- Gemini 4 Carbon: Google's most capable Gemini 4 model, built for long-horizon work.
- Gemini 4 Argon: the first Gemini 4 flagship reasoning model.
- Gemini 3.8 and 3.7 Flash: fast reasoning models in the Gemini 3 Flash line.
- Gemini 3.5 Flash-Lite: low-latency, high-throughput tier.
- Gemini 3.1 Pro Preview: the Gemini 3 Pro-tier reasoning model.
All have a context window of about 1M tokens. For Google image and video, see Nano Banana and Veo.
Start using Gemini API on SandBase
Switch models by changing one field.
- 01
Create an API key
Sign in to the SandBase Console and create a key on the API Keys page. Keep it in an environment variable and never commit it to source control.
TERMINALexport SANDBASE_API_KEY="YOUR_SANDBASE_API_KEY" - 02
Send a chat request
Call the OpenAI-compatible Chat Completions API with model google/gemini-3.8-flash. Any OpenAI SDK works with base URL https://api.sandbase.ai/v1.
TERMINALcurl https://api.sandbase.ai/v1/chat/completions \ -H "Authorization: Bearer $SANDBASE_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model": "google/gemini-3.8-flash", "messages": [{"role": "user", "content": "Explain vector databases in two sentences."}]}' - 03
Connect your AI tools
Use the same key in Claude Code, Codex, Cursor and other clients. See Use in AI tools for setup.
Questions, answered
Gemini 4 Carbon or Argon?
Carbon is positioned above Argon for the most demanding long-horizon tasks. Argon is the first Gemini 4 flagship.
Can Gemini take audio input?
Yes. Current Gemini models on SandBase are tagged for audio input.
Which Gemini is cheapest?
Gemini 3.5 Flash-Lite is the low-latency, high-throughput tier.
Where are older Gemini models?
Gemini 2.5 and earlier 3.x versions are on the Google vendor page.
Build with Gemini API.
One key, consistent APIs, and usage-based pricing across every model on this page.