Models on the gateway
One API Key unlocks every model below—switch anytime in the console with proportional credit conversion.
- deepseek-v4-flash — default tier; blazing-fast coding, chat, and agents
- deepseek-v4-pro — deeper reasoning for complex tasks and long outputs
- deepseek-v3.2 — strong quality at lower cost for high-volume workloads
- glm-5.1 — multilingual and tool-friendly general intelligence
- minimax-m2.7 — long-context and creative generation at scale
Each resource pack includes all models with separate credit quotas. Switch the active model in the console; remaining credits convert proportionally—no silent downgrades.
OpenAI & Anthropic compatible
Drop-in for the tools you already use—change base URL and API key only.
OpenAI protocol: POST /chat/completions · Anthropic protocol: POST /v1/messages
Works with:
- Cline, Continue, Roo Code, Cursor
- Open WebUI and custom agents
- Official OpenAI and Anthropic SDKs
- OpenAI & Anthropic tool calling (streaming SSE passthrough)
- Structured outputs via
response_format/ JSON Schema (forwarded upstream)
Full reference: platform.nanobridge.net/api-docs
Most integrations need only a new base_url and API key—no code rewrite.
Instant Setup via Crypto: Fast, Secure, and Borderless
Resource packs—not surprise invoices. See what you spend as you go.
In the developer console you can:
- Track credit balance and per-model usage
- Review request history and cost breakdown
- Switch billing model and set a default model
- We fully support instant, hassle-free Crypto payments. No card registration required, 100% successful checkout in seconds
- Shared credit pool across all models in a pack
- 30-day validity; quotas shown upfront per tier
- No hidden fees or unexplained balance drops
Email alerts when remaining pack credits fall below 80% (if configured). Balance syncs from call logs about every five minutes.
Global acceleration
Edge nodes in major regions to cut latency and keep routing stable under load.
- United States
- Europe
- Southeast Asia
- India
Latency samples from multiple regions are published for transparency—so developers in the Americas, Europe, and Asia get a predictable experience.
Privacy & data handling
We do not persist prompts, conversations, source code, or uploads for product analytics.
Traffic is processed in transit for routing and inference only.
Customer data is not used to train models.
Read our full Privacy Policy.
Documentation & support
English-first docs and email support for builders worldwide.
- API reference & quick start
- Integration examples for OpenAI and Anthropic clients
- Billing and model-switching guides in the console
- Email support for integration, billing, and routing questions
Open the platform to create keys, buy a pack, and call the API in minutes.
Resource packs
Three tiers, 30 days each, USD pricing. Every pack includes all models with dedicated credit quotas—see deepseek-v4-flash credits per tier below.
Contact
Reach us for partnerships, billing, or technical questions.