OCI
Oracle Cloud Infrastructure is Oracle's cloud. It sells models per token and on dedicated AI clusters through OCI Generative AI, and rents bare-metal and virtual GPU servers by the hour.
- Website: https://www.oracle.com/cloud/
- Headquarters: Austin, United States
- Kind: Hyperscaler
OCI sells 7 of the compared models per token and prices 11 kinds of GPU for hosting your own.
Models sold per token
| Model | Input, per 1M tokens | Output, per 1M tokens |
|---|---|---|
| Cohere Command A | $6.24 | $6.24 |
| Llama 3.1 405B | $10.68 | $10.68 |
| Llama 3.3 70B | $0.72 | $0.72 |
| Llama 4 Maverick 17B | $0.72 | $0.72 |
| Llama 4 Scout 17B | $0.72 | $0.72 |
| OpenAI gpt-oss 120B | $0.15 | $0.60 |
| OpenAI gpt-oss 20B | $0.07 | $0.30 |
GPUs for hosting your own
| GPU | From, per GPU-hour | Cheapest instance | GPUs in it |
|---|---|---|---|
| A10 | $2.00 | VM.GPU.A10.1 | 1 × 24 GB |
| L40S | $3.50 | BM.GPU.L40S.4 | 4 × 48 GB |
| A100 | $4.00 | BM.GPU.A100-v2.8 | 8 × 80 GB |
| MI300X | $6.00 | BM.GPU.MI300X.8 | 8 × 192 GB |
| MI355X | $8.60 | BM.GPU.MI355X.8 | 8 × 288 GB |
| H100 | $10.00 | BM.GPU.H100.8 | 8 × 80 GB |
| H200 | $10.00 | BM.GPU.H200.8 | 8 × 141 GB |
| B200 | $14.00 | BM.GPU.B200.8 | 8 × 180 GB |
| B300 | $15.00 | BM.GPU.B300.8 | 8 × 263 GB |
| GB200 | $16.00 | BM.GPU.GB200.4 | 4 × 192 GB |
| GB300 | $18.00 | BM.GPU.GB300.4 | 4 × 278 GB |
Where it runs
Its models run on demand in 9 regions in 7 countries, per Oracle's documentation; Oracle's price list has one price for every region.
| Country | Regions (Oracle's documentation) | Models on demand |
|---|---|---|
| United States | US Midwest (Chicago) | 6 |
| Germany | Germany Central (Frankfurt) · EU Sovereign Central (Frankfurt) | 4 · 1 |
| Japan | Japan Central (Osaka) | 4 |
| Brazil | Brazil East (Sao Paulo) | 2 |
| United Kingdom | UK South (London) · UK Gov South (London) | 2 · 1 |
| India | India South (Hyderabad) | 1 |
| Saudi Arabia | Saudi Arabia Central (Riyadh) | 1 |
No place stated for 1 model: one global price, and OCI does not say where a request is processed.
Where the prices come from
- Per-token prices: Oracle's price list API
- GPU servers: Oracle's price list API
- Reserved capacity: Oracle's price list API: dedicated AI clusters
- Data sent out (egress): apexapps.oracle.com/pls/apex/cetools/api/v1/products/ (Outbound Data Transfer)
Sources
List prices collected 4 November 2026, refreshed every three hours. No negotiated rates or taxes.