# GPTunneL

> GPTunneL is an AI superapp: one account and one balance give access to 200+ AI models — text, images, video, music, 3D — plus ready-made AI products built on top of them and our own Grom model family. Pay-as-you-go, no subscriptions: you top up once and spend on anything.

## Who we are

- A superapp, not a model aggregator: alongside raw model access you get finished products — assistants, visual AI pipelines, a website chat widget, a page builder, content tools.
- One balance for everything. No per-model subscriptions, no monthly fees, no mandatory payments: if you don't use it, you don't pay.
- Our own Grom models — text, images, video — researched, trained and served on our own hardware, at a fraction of the price of third-party models.
- A single API for every model in the catalog, so the same balance powers your own products.
- Built by ScriptHeads. 2M+ users, 100M+ AI requests processed, 5K+ business accounts.

## Products

- [Prices](https://www.gptunnel.ru/en/prices): Pay as you go, no subscriptions. Up-to-date prices for every model: text per token, images per item, video and music per generation.
- [AI assistants](https://www.gptunnel.ru/en/assistants): Custom AI assistants: your own instructions, knowledge base, MCP and API support, any top model under the hood. A gallery of ready-made assistants plus business assistants for teams.
- [Workflow](https://www.gptunnel.ru/en/workflow): Build a scenario from blocks on a canvas: inputs, LLM, image and video generation, search, memory, voiceover. Describe the task in plain language — the AI assistant builds and runs the workflow.
- [CodeBlocks](https://www.gptunnel.ru/en/codeblocks): Describe it in words — CodeBlocks builds a live HTML page: a landing page, presentation, browser game, calculator or banner. Edit by chatting, share by link, no code and no developers.
- [ChatDeck](https://www.gptunnel.ru/en/chatdeck): Free live chat for your website with an AI agent: it answers first, writes like a person and hands the hard questions to your team. One line of code, pay only for model tokens.
- [AI tools](https://www.gptunnel.ru/en/tools): GPTunneL AI tools: background removal, face swap, text-to-speech, photo enhancement and more.
- [API](https://www.gptunnel.ru/en/docs): A single API to every GPTunneL AI model. Add text, images, video and audio to your products.
- [For business](https://www.gptunnel.ru/en/b2b): Corporate access to 200+ AI models: a single company account, monthly invoicing, role-based access and per-employee spend analytics.
- [AI consulting](https://www.gptunnel.ru/en/b2b/consulting): Process audit, a map of use cases, a pilot in two weeks and team training: we embed AI into your workflows on top of GPTunneL, turnkey.
- [Gift cards](https://www.gptunnel.ru/en/gift-cards): GiftCard — virtual GPTunneL gift cards: a promo code loaded with bonuses, any amount from 100 bonuses, personal message. For friends, teams and giveaways.
- [Partnership & White Label](https://www.gptunnel.ru/en/partnership): Two ways to work with GPTunneL: a referral program that pays you a share of what your users spend, and white label — our service under your brand via an agency agreement.
- [Web app](https://app.gptunnel.ai)

## Our own models (Grom)

We research, train and serve these ourselves — which is why they cost a fraction of comparable third-party models:

| Model | Price |
|---|---|
| Grom 1.5 Pro⚡️ | $0.1 / $0.4 per 1M tokens (input / output) |
| Grom Art | from $0.04 per image |
| Grom Nova | from $0.005 per image |
| Grom Pixel | $0.16 per image |
| Uzor | from $0.16 per image |
| Uzor Lite | $0.04 per image |
| Grom TV | from $0.05 per second |

## How pricing works

- Pay-as-you-go: you top up a single balance and spend it on any model. No subscriptions, no mandatory payments.
- Text is billed per token, images per image, video per second of footage, music per track.
- The same balance covers the web app, the API and every product below.

Loyalty discount, applied to every model in the catalog and based on your spend over the last 30 days:

Plus 5% (from $5) · Bronze 8% (from $20) · Silver 12% (from $60) · Gold 17% (from $120) · Platinum 22% (from $300) · Diamond 25% (from $600)

## What things actually cost

Real numbers from the current catalog:

- 1M input + 1M output tokens on Claude Opus-5 — $90
- one image: Nano Banana Pro — $0.15, our Grom Nova — $0.005
- a five-second clip: Kling 2.6 — $0.71, our Grom TV — from $0.25
- a track with vocals on Suno — $0.12

## Popular models and prices

### Text

| Model | Input / output per 1M tokens | Context |
|---|---|---|
| Claude Fable 5 | $20 / $100 | 1M |
| Claude Opus-5 | $15 / $75 | 1M |
| GPT 5.6 Sol | $10 / $60 | 1.05M |
| Grok 4.5 | $4 / $12 | 500K |
| Gemini 3.1 Pro | $6 / $30 | 1.05M |
| DeepSeek V4 Pro | $0.88 / $1.74 | 1.05M |
| Kimi K3 | $6 / $30 | 1.05M |

### Images

| Model | Price |
|---|---|
| Grom Art | from $0.04 per image |
| Grom Nova | from $0.005 per image |
| Nano Banana 2 | from $0.16 per image |
| Nano Banana 2 Lite | $0.096 per image |
| Image 2 | from $0.32 per image |
| Seedream 5 Pro | from $0.09 per image |

### Video

| Model | Price |
|---|---|
| FLUX 3 | from $0.12 per second |
| Grom TV | from $0.05 per second |
| Grok Imagine 1.5 | from $0.14 per second |
| Seedance 2 | from $0.28 per second |
| Kling v3 | from $0.34 per second |
| WAN 2.7 | from $0.2 per second |
| Gemini omni | from $0.23 per second |

### Music

| Model | Price |
|---|---|
| MiniMax Music 3.0 | $0.15 per track |
| Mureka | $0.09 per track |
| Suno | $0.12 per track |

Prices are indicative and taken from the public catalog snapshot; the live per-model price list lives in the app. https://app.gptunnel.ai/prices

## Model reference pages

Plain-language overviews of what each model does, what it costs and when to pick it:

- [ChatGPT without a VPN — access from anywhere](https://www.gptunnel.ru/en/model/chatgpt): Every OpenAI model in GPTunneL: the GPT-5 family, Codex for code and the o-series. Pay per token, no subscription and no VPN.
- [The Claude family — general-purpose LLMs by Anthropic](https://www.gptunnel.ru/en/model/claudel-ot-anthropic): All Claude models in GPTunneL: Opus, Sonnet, Haiku and Fable. Context up to 1M tokens, pay per token, no subscription and no VPN.
- [Gemini by Google: a million tokens and PDF reading](https://www.gptunnel.ru/en/model/gemini): Google Gemini models in GPTunneL: 1M token context, PDF and image understanding, open Gemma weights. Pay per token, no subscription and no VPN.
- [Grok by xAI: a 2M token context and live web search](https://www.gptunnel.ru/en/model/grok): Grok models by xAI in GPTunneL: context up to 2M tokens, web search and image generation. Pay per token, no subscription and no VPN.
- [DeepSeek — frontier quality at a fraction of the price](https://www.gptunnel.ru/en/model/deepseek): DeepSeek models in GPTunneL: 1M token context, a reasoning mode and open weights. Pay per token, no subscription and no VPN.
- [Nano Banana — edit an image in plain words](https://www.gptunnel.ru/en/model/nano-banana): Google Nano Banana models in GPTunneL: edit an image in plain words, keep the same character across frames, blend photos. Pay per image, no VPN.
- [GPT Image — OpenAI's image generator, no VPN needed](https://www.gptunnel.ru/en/model/gpt-image): OpenAI GPT Image models in GPTunneL: long prompts followed to the letter, legible text on a layout, edits inside a selection. Pay per image, no VPN.
- [Midjourney without Discord or a subscription](https://www.gptunnel.ru/en/model/midjourney): Midjourney in GPTunneL: its signature look with no monthly subscription and no Discord account. Pay per image from one shared balance.
- [FLUX — an image and video generator with sound](https://www.gptunnel.ru/en/model/flux-ai---generator-izobrazheniy-novogo-pokoleniya): FLUX models by Black Forest Labs in GPTunneL: photorealism, readable text in frame and video generation with sound. Pay per generation, no subscription.
- [Seedream by ByteDance — 4K frames out of the box](https://www.gptunnel.ru/en/model/seedream): ByteDance Seedream models in GPTunneL: 4K output out of the box, generation and editing in one model, legible captions. Pay per image, no VPN.
- [Veo by Google — video with sound in one render](https://www.gptunnel.ru/en/model/veo): Google DeepMind Veo models in GPTunneL: video with sound in the same render, spoken lines in frame, clips from a photo. Pay per second, no VPN.
- [Kling AI: a video generation model, the Chinese answer to Sora](https://www.gptunnel.ru/en/model/kling-ai---neyroset-dlya-generacii-video-po-tekstu): Kling models by Kuaishou in GPTunneL: video from text and from a photo, natural motion, sound in the same render. Pay per second, no subscription.
- [Seedance by ByteDance — video that cuts between shots](https://www.gptunnel.ru/en/model/seedance): ByteDance Seedance models in GPTunneL: several shots in one generation, natural motion, fast rendering. Pay per second, no subscription and no VPN.
- [Suno AI - online music and video generation](https://www.gptunnel.ru/en/model/suno): Suno in GPTunneL: a song with vocals, melody and arrangement in one generation, your own lyrics and your own genre. Pay per generation, no subscription.

## More

- [Blog](https://www.gptunnel.ru/en/blog): Guides, news and ideas for working with AI.
- [About the company](https://www.gptunnel.ru/en/about): ScriptHeads is the team building GPTunneL — AI products crafted with a focus on quality, stability, and reliable partnership.
- [Legal documents](https://www.gptunnel.ru/en/documents): GPTunneL legal documents: terms of service, privacy policy, offer and other official documents.

## Other languages

- This file is also available in Russian: https://www.gptunnel.ru/ru/llms.txt