Introduction
What SoleAPI is, the base URL, and the three API dialects at a glance.
SoleAPI is a multi-model AI gateway: call Claude, GPT, Gemini, Grok, DeepSeek and other leading models through one API key and one stable endpoint. Smart routing and failover, per-token billing, rate limiting and quota governance, request data masking, and real-time usage analytics are built in.
Base URL
All requests go to a single host:
https://api.soleapi.comWith official SDKs, just point base_url at the address above — no other code changes needed.
Three API dialects
The gateway exposes all three API shapes, and one key works for all of them. Pick whichever your existing code or tooling already speaks:
| Dialect | Endpoint | Auth header | Typical clients |
|---|---|---|---|
| OpenAI-compatible | POST /v1/responses | Authorization: Bearer | OpenAI SDK · Codex |
| Anthropic-compatible | POST /v1/messages | x-api-key | Anthropic SDK · Claude Code |
| Gemini-compatible | POST /v1beta/models/{model}:generateContent | x-goog-api-key | Google GenAI SDK |
The gateway translates between protocols where needed, so most chat models can be called from any dialect. If a model only accepts a specific protocol, you get a clear error explaining it (see Errors).
Core capabilities
- Smart routing & failover: automatic routing across multiple upstreams, with failover and circuit-breaker recovery when one goes down.
- Per-token billing: input and output (including cache and reasoning tokens) priced separately, with a full balance and transaction ledger.
- Rate limits & quotas: RPM/concurrency limits, per-key spending limits with optional periodic resets, and IP allowlists.
- Data masking & content compliance: configurable privacy filters mask sensitive fields before forwarding; content moderation blocks risky requests.
- Real-time observability: dashboard trends, per-key usage stats, and per-request detail logs.
Next steps
- Quickstart: make your first call in minutes.
- Authentication: the three auth headers and key management.
- Models: list the models available to your key; see the model catalog for pricing.