Pay for intelligence, not subscriptions.
Every model has a price per token. We show all of them, live, at aquire.cc/pulse
What's your number?
18
18
105
30 chats and 5 coding sessions a day.
Three AI subscriptions: $60 a month.
Aquire with Auto routing: $5.13 a month, with access to every model.
Prices are live. Run your own numbers at aquire.cc/pulse
4
5
60
What does a question to an AI actually cost?
Depends on which model you ask. Across ten popular models, the same 1,000 chats cost anywhere from $0.29 to $15.12.
Aquire routes every request to the cheapest model that can handle it, and you pay per token with ETH. No subscription.
aquire.cc/pulse
Same question. 52x the price.
We priced 1,000 everyday chats across ten popular models, live:
Claude Opus 5.5: $15.12
GPT-5.6 Terra Pro: $8.42
Grok 4.7: $5.83
Llama 4 Maverick: $1.00
DeepSeek V4.1 Flash: $0.29
Most questions don't need the most expensive model. Aquire Auto sends each one to the cheapest model that can do the job.
aquire.cc/pulse
13
11
139
Introducing Aquire Pulse: the live price of intelligence.
Every model on Aquire, priced in real time. 372 models, the live ETH rate, and a calculator that shows what your own AI use would cost per token.
Light use: $0.70 a month.
Everyday use: $5.13 a month.
Power user: $20.30 a month.
Versus $60 for three subscriptions.
Find your number: aquire.cc/pulse
1
1
36
Aquire is built for AI agents.
Agents don't make one AI call. They make thousands, in loops, while nobody's watching. A card and a monthly plan were never designed for that.
Here's how it works on Aquire.
One balance, funded with ETH. Top up once from your wallet. Your ETH converts to dollar credits at the live rate. No card, no KYC, no subscription, and credits never expire.
A key per agent. Every agent gets its own API key, with a hard spend cap and its own list of allowed models. Pause or revoke any key instantly. Up to 20 keys per account.
400+ models, one API. GPT, Claude, Gemini, DeepSeek, Llama and more behind one OpenAI-compatible endpoint. If your agent runs on the OpenAI SDK, it runs on Aquire. Change the base URL and you're done.
Auto routing for loops. Set the model to aquire/auto and every call goes to the cheapest model that can handle it. Simple steps run on fast, cheap models. Hard reasoning still gets a frontier model.
No surprise bills. When a key hits its cap or the balance runs out, calls stop with a clear error your agent can read. You never spend more than you put in.
MCP ready. Connect Claude Desktop, Cursor and other MCP clients with a single key.
Next: agent wallets, so agents can hold their own balance and pay per request on-chain.
aquire.cc/docs
6
7
83
You get 400+ AI models, one key, and you only pay for what you use, in ETH.
What do you build first? 👇
2
5
68
How paying with ETH works on Aquire 🔍
1. You get an exact amount
2. You send it from your wallet
3. We verify it on-chain: right wallet, exact amount, success, confirmed
4. Credits land in seconds
No card. No middleman. Each transaction can only be used once.
6
9
126
Phase 1 is live. Here's what's next 👇
📱 Mobile wallet sign-in
💳 Card, email and Google
🤝 Partner program for Robinhood Chain apps
🤖 Agent wallets with spend limits
Building in public: aquire.cc/roadmap
2
5
43
Builders: change one line, get 400+ models.
Aquire speaks the OpenAI API. Point base_url at aquire.cc/v1 and keep your code. Works with LangChain, Vercel AI SDK, LiteLLM, and MCP clients like Claude Desktop and Cursor.
Get a key 👉 aquire.cc/app
1
13
Aquire is built for AI agents.
Agents don’t make one AI call. They make thousands, in loops, while nobody’s watching. A card and a monthly plan were never designed for that.
Here’s how it works on Aquire.
One balance, funded with ETH. Top up once from your wallet. Your ETH converts to dollar credits at the live rate. No card, no KYC, no subscription, and credits never expire.
A key per agent. Every agent gets its own API key, with a hard spend cap and its own list of allowed models. Pause or revoke any key instantly. Up to 20 keys per account.
400+ models, one API. GPT, Claude, Gemini, DeepSeek, Llama and more behind one OpenAI-compatible endpoint. If your agent runs on the OpenAI SDK, it runs on Aquire. Change the base URL and you’re done.
Auto routing for loops. Set the model to aquire/auto and every call goes to the cheapest model that can handle it. Simple steps run on fast, cheap models. Hard reasoning still gets a frontier model.
No surprise bills. When a key hits its cap or the balance runs out, calls stop with a clear error your agent can read. You never spend more than you put in.
MCP ready. Connect Claude Desktop, Cursor and other MCP clients with a single key.
Next: agent wallets, so agents can hold their own balance and pay per request on-chain.
aquire.cc/docs
ETH isn’t used per token. When you top up, your ETH converts to USD credits at the live ETH/USD rate (avg of Coinbase + Kraken) at that moment. After that everything is priced in dollars.
Each request = the model’s live per-token price (input + output tokens, same as the provider charges) + a flat 8%. Every model’s price is public at aquire.cc/v1/models
Rate limit: 120 requests/min per account, across all keys. Up to 20 API keys per account, each with its own spend cap.
if you pay with eth, how do you calculate cost per token across 400+ models and what is the rate limit per account
4
5
145
Same AI. Up to 90% cheaper.
Auto routing sends every request to the cheapest model that can do the job well. Hard problems still get a frontier model.
Try it: aquire.cc/app
4
113
Aquire retweeted
You're paying $60 a month for 3 AI apps.
Aquire: one key, 400+ models. Pay only for what you use, with ETH from your wallet.
Next: your portfolio's interest pays for it.
aquire.cc/app
@AquireCC
You're paying $60 a month for 3 AI apps.
Aquire: one key, 400+ models. Pay only for what you use, with ETH from your wallet.
Next: your portfolio's interest pays for it.
aquire.cc/app
@AquireCC