One API for every LLM.
Automatically routed for lower cost.
One Basktre API key unlocks a growing catalog spanning GPT, Claude, Gemini, Llama, DeepSeek, and more. Live availability may vary during beta. Set model: "auto" and we route each request to the cheapest model that can do the job. Cost-aware model selection, automatically. We charge a 4% top-up fee. The remaining wallet balance is used entirely for model costs.
No prompts stored. · No response logs. · 4% platform fee. · That's it.
curl https://api.basktre.in/api/v1/models/stream \ -H "api-key: your_basktre_key" \ -H "Content-Type: application/json" \ -d '{ "model": "auto", "messages": [{"role": "user", "content": "Explain AI gateway"}] }'
Your first $1.00 of AI calls
is on us. seriously.
Create a free Basktre account and receive $1 USD in instant wallet credits — no credit card, no trial period, no strings. Use them across currently available AI models at displayed per-token rates with no additional Basktre token markup.
- No credit card required
- Instant credit, no waiting
- Works across the live model catalog
- No expiry on credits
Free account · No credit card · $1 lands in your wallet immediately
You say "auto".
We do the math.
Every request is analysed — length, complexity, task type — and routed to the cheapest model that can actually do the job. No config, no prompt engineering. Just set model: "auto" and go home early.
task: Summarise a 2-page PDF.
tokens: ~320 tokens
task: Multi-step legal analysis.
tokens: ~4 800 tokens
task: Write a React data table.
tokens: ~1 200 tokens
Want a specific model? model: "gpt-4o" works exactly as you'd expect. Auto is opt-in, not mandatory.
Everything you need.
Nothing you don't.
Stop juggling 6 dashboards
One Basktre key unlocks every major model. Swap providers without touching your code.
Let the router earn its keep
Send model: "auto" and we pick a cost-effective model that can handle the task. Pin a model whenever consistency matters.
Full control when you want it
Need GPT-4o for a specific task? model: "gpt-4o" — no lock-in, override anything.
Your prompts are none of our business
Prompt and response bodies are not stored or logged. Limited usage metadata remains for billing and operations.
Top up. Use. Done.
Add funds to your wallet. We charge a 4% top-up fee, and the remaining balance is used entirely for model costs.
Every frontier model, one place
Major model families in one place, with pricing and availability regularly updated.
Every model.
Transparent pricing.
Pricing shows the displayed per-token model rates with no additional Basktre token markup. The only fee is 4% when you top up your wallet. Browse, compare, and pick the right model for every task.
| Model▲ | Input / 1M▲ | Output / 1M▲ | Total / 1M▲ | Tier |
|---|---|---|---|---|
AN Claude Haiku 4.5 | $1 | $5 | $6 | Mid |
AN Claude Opus 4.6 | $5 | $25 | $30 | Pro |
AN Claude Opus 4.7 | $5 | $25 | $30 | Pro |
AN Claude Sonnet 4.6 | $3 | $15 | $18 | Mid |
DS DeepSeek-V4-Flash | $0.14 | $0.28 | $0.42 | Nano |
DS DeepSeek-V4-Pro | $2 | $3 | $5 | Mid |
OA GPT-4.1 gpt-4.1 | $2 | $8 | $10 | Mid |
OA GPT-4.1 Mini gpt-4.1-mini | $0.40 | $2 | $2 | Budget |
OA GPT-4o | $3 | $10 | $13 | Mid |
OA GPT-5 | $1 | $10 | $11 | Mid |
We see the call.
Never the content.
Basktre is infrastructure. We route your requests — we don't read them, log them, or use them for anything. These aren't policies. They're architectural constraints.
Zero logs
Your prompts never touch our database. Requests are proxied through and immediately discarded — we have no access to your conversation content.
Zero training
We don't use your data to improve models, fine-tune anything, or sell insights. Not now, not ever. It's not our data.
Zero markup
4% on wallet top-ups. The remaining balance is used for model costs at displayed per-token rates, with no additional Basktre token markup.
Frequently asked.
Ready to simplify
your AI stack?
Sign up free. Top up when you're ready. 4% fee. No subscriptions, no seat costs, no surprises.
No prompts stored. No response logs. 4% platform fee. That's it.