Google Gemini API
Google builds the Gemini model family through Google DeepMind and serves it through the Gemini app, Google AI Studio, the Gemini API and Vertex AI. Its portfolio spans frontier Gemini reasoning models, Flash-tier models, Nano Banana image models, Veo video models and open-weight Gemma models. Through Felo, eight Google routes are available with one Felo key and one balance, with no Google Cloud project or API key to set up.
- Models
- 15
- chat ×8 · image ×3 · video ×3 · decisions ×1
- Input per 1M
- $0–$4
- Output per 1M
- $0–$20
- Largest context
- 1.049M
What Google provides
Google DeepMind develops the Gemini models, and Google exposes them through several surfaces: the Gemini app for end users, AI Studio for prototyping, the Gemini API for developers and Vertex AI for Google Cloud deployments. The Gemini API accepts text, image, audio, video and PDF input on the larger models and supports search grounding, code execution, function calling, context caching and a Batch API. Gemini image generation is branded Nano Banana, video generation is Veo, and Gemma covers open-weight models that can run outside Google's cloud.
Google models available on Felo
Felo currently lists eight Google routes: Gemini 4 Argon, Gemini 3.8 Flash, Gemini 2.5 Pro, Nano Banana Pro (Gemini 3 Pro Image), Nano Banana 2.1, Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image), Veo 3.1 Fast and Gemma 4 26B A4B. They cover text and multimodal chat, image generation and editing, and video generation; the live table on this page is the authoritative list with current rates and limits.
Pricing and access
On Felo, Google usage is billed from your Felo balance at the rate published for each route. Published default-channel rates for the Gemini chat models currently range from $0 for the Gemma 4 route to $4 per 1M input tokens and $20 per 1M output tokens for Gemini 4 Argon, with the Flash and 2.5 Pro tiers in between. Google prices Nano Banana output per generated image and Veo output per second, and Felo's model table shows the published unit and rate for each route. Google's direct Gemini API has a free tier for some models and a 50% Batch discount, but those terms apply to Google's platform, not to Felo usage. No Google Cloud account, project or API key is needed to call these models through Felo.
Why teams use Google models through Felo
Gemini 4 Argon is built for long-horizon software engineering, enterprise knowledge work and cybersecurity defense, while Gemini 3.8 Flash targets agents and workflows where latency and cost matter more than the top tier. Nano Banana and Veo add image and video generation, so one application can cover text, image and video without integrating three separate Google surfaces. In Felo, those routes sit next to other providers on the same key, which makes cross-provider comparison and fallback straightforward.
Limits and access notes
Google releases its newest frontier models in stages: Gemini 4 Argon first went to trusted cyber defenders through the Fairwind Program and then to paid API customers and Google AI Ultra subscribers, so availability can lag the announcement. Not every Gemini feature is exposed on every Felo route — the catalog notes that some Gemini 3.x and 4 routes do not support the Live API, audio generation or explicit caching — and the model page is the place to confirm supported modalities and protocols. Gemini 2.5 Pro remains listed as an older option, and the image and video routes have their own output limits rather than the chat models' context windows.
- Input per 1M
- $1.5
- Output per 1M
- $7.5
- Context
- 1.049M
- Max output
- 943.718K
- Input per 1M
- $0.75
- Output per 1M
- $3.75
- Context
- 1.049M
- Max output
- 943.718K
- Input per 1M
- $0.1
- Output per 1M
- $0
- Context
- 262.144K
- Max output
- 235.929K
- Input per 1M
- $4
- Output per 1M
- $20
- Context
- 1.049M
- Max output
- 943.718K
- Input per 1M
- $2
- Output per 1M
- $10
- Context
- 1.049M
- Max output
- 943.718K
- Input per 1M
- Not published
- Output per 1M
- Not published
- Context
- 131.072K
- Max output
- 65.536K
- Input per 1M
- Not published
- Output per 1M
- Not published
- Context
- 131.072K
- Max output
- 65.536K
- Input per 1M
- $4
- Output per 1M
- $20
- Context
- 1.049M
- Max output
- 943.718K
- Input per 1M
- Not published
- Output per 1M
- Not published
- Context
- 65.536K
- Max output
- 58.982K
- Input per 1M
- $1.5
- Output per 1M
- $7.5
- Context
- 1.049M
- Max output
- 65.536K
- Input per 1M
- $0.25
- Output per 1M
- $1.5
- Context
- 65.536K
- Max output
- 58.982K
- Input per 1M
- $2
- Output per 1M
- $12
- Context
- 131.072K
- Max output
- 32.768K
- Input per 1M
- Not published
- Output per 1M
- Not published
- Context
- —
- Max output
- —
- Input per 1M
- $0
- Output per 1M
- $0
- Context
- 262.144K
- Max output
- 32.768K
- Input per 1M
- $1.25
- Output per 1M
- $10
- Context
- 1.049M
- Max output
- 65.536K
| Model | Input per 1M | Output per 1M | Context | Max output |
|---|---|---|---|---|
| $1.5 | $7.5 | 1.049M | 943.718K | |
| $0.75 | $3.75 | 1.049M | 943.718K | |
| $0.1 | $0 | 262.144K | 235.929K | |
| $4 | $20 | 1.049M | 943.718K | |
| $2 | $10 | 1.049M | 943.718K | |
| Not published | Not published | 131.072K | 65.536K | |
| Not published | Not published | 131.072K | 65.536K | |
| $4 | $20 | 1.049M | 943.718K | |
| Not published | Not published | 65.536K | 58.982K | |
| $1.5 | $7.5 | 1.049M | 65.536K | |
| $0.25 | $1.5 | 65.536K | 58.982K | |
| $2 | $12 | 131.072K | 32.768K | |
| Not published | Not published | — | — | |
| $0 | $0 | 262.144K | 32.768K | |
| $1.25 | $10 | 1.049M | 65.536K |
Prices are published standard rates in USD per 1M tokens, before channel discounts.
Frequently asked questions
- What is the Google Gemini API?
- It is Google's developer API for the Gemini family, served through Google AI Studio and Vertex AI. It supports text and multimodal input, grounding and tool features, image generation through Nano Banana and video generation through Veo.
- Which Google models can I call through Felo?
- Felo lists eight Google routes: Gemini 4 Argon, Gemini 3.8 Flash, Gemini 2.5 Pro, three Nano Banana image models, Veo 3.1 Fast and Gemma 4 26B A4B. The live table on this page is the authoritative list.
- Do I need a Google Cloud or AI Studio key?
- No. Google models on Felo use a Felo API key and are billed from your Felo balance; you do not need a Google Cloud project, an AI Studio key or a Google API key.
- How is Google usage billed on Felo?
- Per the unit published for each route: tokens for Gemini chat models, generated images for Nano Banana and output seconds for Veo. The model table shows the current rate, and Google's direct-API free tier and Batch discount do not apply to Felo usage.
- Is Gemini 4 Argon available through Felo?
- Yes. Felo lists Gemini 4 Argon as a chat route; Google released it in stages, first to cyber defenders and then to paid API customers and AI Ultra subscribers, and it carries Google's post-introductory rate of $4 per 1M input and $20 per 1M output tokens.
- Is there a free Google model on Felo?
- Yes. Gemma 4 26B A4B is listed with a published rate of $0 on Felo, while the Gemini chat, image and video routes are paid. The live table is the authoritative check for current rates.