| What it does | Multi-protocol AI router for Claude, GPT, Gemini and 100+ models with automatic failover at 15% below official pricing
| A scheduling API for developers and their agents to make HTTP calls later, with automatic retries and full record-keeping
|
|---|
| Stage | Not given
| Not given
|
|---|
| Founded | Not given
| Not given
|
|---|
| Based in | US
| US
|
|---|
| Pricing model | Paid
| Paid
|
|---|
| Pricing | No subscription. Pay per token at 85% of official list price. Start free with $1 credit.
| $1 per 100,000 attempts. First project included in subscription. Additional projects cost $5/month. Trial credit of $1 valid for 14 days upon sign-up. Prepaid credit from $10, never expires.
|
|---|
| Key features | - One API key for 100+ frontier models
- Automatic failover across providers
- OpenAI-compatible API
- Transparent per-token pricing
- Live usage dashboard
- Native Anthropic, OpenAI, Gemini support
- Works with Claude Code, OpenCode, Codex CLI
- Zero data retention
- Per-key usage logs and limits
| - Schedule calls at a specified delay, at an exact time, on a cron expression, or at regular intervals
- Automatic retries with growing waits, up to 5 failures
- Full attempt records with status, timing, and response body up to 10 KiB
- Cryptographic signing of all calls verified over exact bytes
- Time zone support with daylight saving handling
- Respects HTTP Retry-After headers, waiting up to one hour
- Built-in MCP server for AI agents like Claude Code, Codex, and Cursor
- Bulk operations to manage many schedules together with dashboard, API, CLI, or SDK access
- REST API, JavaScript SDK, CLI, and dashboard interfaces
- 24-hour forecast for autoscalers
|
|---|
| What makes it different | - 15% cheaper than official list prices for all models
- No vendor lock-in with OpenAI-compatible code
- Automatic failover in milliseconds
- Zero data retention and privacy
- Set up in 2 minutes with one base URL change
| - Built from the start for AI agents through Model Context Protocol
- Pay only for attempts made; waiting costs nothing
- Automatic retry handling removes need to build job queue infrastructure
- Every attempt recorded with full context for debugging
- Cryptographically signed calls allow verification before trusting
|
|---|
| Community votes | 0 | 0 |
|---|
| DR (Domain Rating by Ahrefs) | 16 ▼ −1 (down) since last check | 2 ▲ +2 (up) since last check |
|---|
| Trust Flow (Majestic) | 12 | 2 |
|---|
| Citation Flow (Majestic) | 26 | 23 |
|---|
| Referring domains (Majestic) | 79 | 10 |
|---|
ApiFlux
GreatWait