GPT Open Source API Pricing
4 available providers and 4 live price sources tracked, including 0 official provider.
Data snapshot:
30-second decision
- Lowest recorded input price: $0.0500/1M tokens
- Compare official API and relay quotes with their service terms
Provider summary and buying decision
Evidence tier first, then normalized price within each tier.
| Provider / operator | Price evidence | Public operating history | Input / 1M tokens | Output / 1M tokens | API docs | Refund policy | Payment / invoice |
|---|---|---|---|---|---|---|---|
| Together AIunknown | Live API pricePlatform page ↗Observed: | —Unknown | $0.05 | $0.20 | View docs ↗ | Not provided | visa · mastercardunknown |
| Fireworks AIunknown | Live API pricePlatform page ↗Observed: | —Unknown | $0.07 | $0.30 | View docs ↗ | Not provided | visa · mastercardunknown |
| OpenRouterunknown | Live API pricePlatform page ↗Observed: | —Unknown | $0.075 | $0.30 | View docs ↗ | Not provided | visa · mastercard · cryptounknown |
| OhMyGPTunknown | Live API pricePlatform page ↗Observed: | —Unknown | $0.077 | $0.33 | View docs ↗ | Not provided | alipay · wechatunknown |
Platform-submitted information or domain verification does not mean the price was independently checked and is not a ComputeUnion guarantee of pricing, availability, refunds, compliance, invoices, or service quality.
Estimate monthly cost
The estimate uses the displayed price without assuming discounts or cache hits.
External capability references
Only source results safely mapped to the official model name are shown.
FAQ
How is GPT Open Source API priced?
Input and output tokens are billed separately. A complete official input/output baseline is not currently recorded, so provider quotes are normalized per million tokens.
Which provider is cheapest for GPT Open Source?
The lowest recorded input price is $0.0500/1M tokens from Together AI; verify provider terms and the observation date before production use.
How do official and relay prices differ?
Official and relay channels can differ in price, latency, limits, and service terms; evaluate them separately for production traffic.