1 models from 1 makers that you can call through Anthropic Claude API — each one tested by us with fixed prompts, outputs kept, costs measured.
Makers: anthropic
Docs Get an API key Their pricing page
The Claude API is Anthropic's direct, first-party channel for the Claude model family — Opus, Sonnet, and Haiku — accessed through the Anthropic Console.
Current frontier models support extended thinking (an explicit reasoning mode) and large context windows (200K+ tokens). Anthropic doesn't run a public self-serve affiliate program for API signups; its referral programs are either enterprise consultancy deals or a small consumer credit perk on Claude Max, unrelated to API usage.
Affiliate program: Anthropic runs an approval-gated enterprise “Referral Partner Program” for consultancies (an undisclosed one-time fee per referred customer) and a separate Claude Max consumer credit perk ($10 per referral, capped at $30) — neither is a public, self-serve API affiliate program.
This is what Anthropic Claude API itself charges for each model — not the cheapest price across every provider (that's what the cards above show).
| Model | Maker | Price via Anthropic Claude API | Context |
|---|---|---|---|
| Anthropic: Claude 3 Haiku | anthropic | — | 200K |
Latency: Per-provider latency isn't independently measured yet — the probe rig currently tests one route per model, not one per provider (see PROBE_RIG.md ledger #10).