Provider comparisons
How this gateway compares
Every comparison below is built from the other provider’s own published prices, read on 2026-10-26 and cited on each page. Where a competitor is genuinely the better choice, the page says so — a comparison you cannot trust is worse than none.
vs OpenRouter
A very large model marketplace with automatic provider fallbacks.
- Models listed
- 467
- Shared with us
- 4
- Cheapest shared
- $0.04 vs $0.30
vs DeepInfra
Low-cost inference on its own GPU fleet, with aggressive open-weight pricing.
- Models listed
- 175
- Shared with us
- 9
- Cheapest shared
- $0.04 vs $0.20
vs Together AI
Serverless and dedicated inference on open-weight models, plus fine-tuning.
- Models listed
- 92
- Shared with us
- 6
- Cheapest shared
- $0.04 vs $0.30
What every comparison covers
Per-model rate cards
The same model id priced at both providers, split into input, cached and output, so a difference in one column is not hidden by an average.
Modelled monthly bills
Four realistic traffic shapes — chatbot, coding agent, RAG and batch — converted into a monthly number at each rate card.
Commercial terms
Platform fees, credit-purchase surcharges, free tiers and funding methods, in plain words rather than marketing copy.
Honest strengths
A section naming where the other provider genuinely wins, because a comparison with no downsides is an advertisement.
Also worth reading
Cheapest model here: DeepSeek V4.1 Flash at $0.04/1M input
Start on the cheaper rate card
Generate a key, point your existing SDK at the base URL, and send a request. There is no email, no card and no KYC step.