Oracle OCI Generative AI
- Categories
- Hyperscaler AI platforms
- Owned by
- Oracle Corporation (NYSE: ORCL)
- HQ
- US
- Website
- https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/
- Pricing page
- https://www.oracle.com/cloud/price-list/
- Fee model
- per_token+per_second_gpu
- Fees
- On-demand: newer models (OpenAI gpt-oss, xAI Grok, Google Gemini) per 1M tokens; Meta Llama and Cohere billed per 10,000 "transactions" where 1 transaction = 1 character of prompt+response (e.g. Llama 4 Maverick/Scout $0.0018 per 10k chars). Dedicated AI clusters billed per AI-unit-hour (e.g. Large Meta dedicated $12/unit-hour).
- Free tier
- no
- OpenAI-compatible API
- partial
- Bring your own key
- no
- Notes
- Prices from Oracle cost-estimator price API (PAY_AS_YOU_GO, USD); the price-list page renders unit prices client-side. Per-character Llama pricing is not convertible to per-token without an assumed chars/token ratio, so those rows leave price cells empty.
- Sources
- https://www.oracle.com/cloud/price-list/ https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/?currencyCode=USD collected 2026-09-18
LLM prices (11 models, USD per 1M tokens)
| Model | Input | Output | Cached input | Context | Source |
|---|---|---|---|---|---|
| Gemini 2.5 Flash | $0.30 | $2.50 | — | — | ✓ src src2 |
| Gemini 2.5 Flash-Lite | $0.10 | $0.40 | — | — | ✓ src src2 |
| Gemini 2.5 Pro | $1.25 | $10 | — | — | ✓ src src2 |
| gpt-oss-120b openai.gpt-oss-120b |
$0.15 | $0.60 | — | — | ✓ src src2 |
| gpt-oss-20b openai.gpt-oss-20b |
$0.07 | $0.30 | — | — | ✓ src src2 |
| Grok 4.20 (Reasoning / Non-Reasoning) | $1.25 | $2.50 | $0.20 | — | ✓ src src2 |
| Grok 4.3 | $1.25 | $2.50 | $0.20 | — | ✓ src src2 |
| Grok 4.6 | $2 | $6 | $0.50 | — | ✓ src src2 |
| Llama 3.1 405B Instruct | — | — | — | — | ◐ src src2 |
| Llama 4 Maverick 17B Instruct | — | — | — | — | ◐ src src2 |
| Llama 4 Scout 17B Instruct | — | — | — | — | ◐ src src2 |