Our take
OpenAI is a US-headquartered provider of hosted-only models with 86 tracked offers, the broadest range of price points in its own lineup, and the highest top-end throughput among its own data. It holds an independent security attestation, does not train on prompts, and hosts an EU-region endpoint.
Use this for maximum model breadth from a single provider. Pick it for latency-sensitive workloads: GPT-5 Mini at 137 tokens per second is the fastest in this provider's data. Choose it for a budget entry point: GPT-5 Nano is the cheapest tracked offer here. Skip it if you need a verified zero-retention option, or consistent throughput across identically-priced variants.
- 86 tracked proprietary offers — the broadest range of price points in this provider's own data.
- Highest throughput among its own tracked offers: GPT-5 Mini at 137 tokens per second, 29% faster than GPT-5 Nano and 315% faster than GPT-4o Mini.
- Cheapest entry point in its own lineup: GPT-5 Nano, 50% cheaper than GPT-5-4 Nano and 83% cheaper than GPT-4o Mini.
- Strong compliance posture: SOC 2 attestation, no training on prompts, and an EU-region endpoint.
- Same model with inconsistent throughput at identical pricing: GPT-4o Mini at 33 tokens per second versus an identically-priced 2024-07-18 variant at 66 tokens per second — a 100% throughput difference with no price distinction.
- Throughput unverified for the search-preview variant of GPT-4o Mini.
- Zero-retention option is unverified in our data.
Privacy & data handling
- ✓
- Yes
- ✗
- No
- No answer on record
The colour says whether the answer favours you, not whether it is a yes — not training on your prompts earns a green cross, no zero-retention option earns a red one.