Dolphin LLM vs the OpenAI API
Updated · prices checked on that date
Both speak the same API format, so switching is a matter of changing the base URL, the key and the model name. The differences are in price, context length, the model itself and how you pay.
Key points
- OpenAI GPT-6 Luna costs $0.10 per 1M input and $0.50 per 1M output tokens; Dolphin LLM costs $0.15 and $0.50.
- OpenAI's models have a far longer context window (1,050,000 tokens) than Dolphin LLM (64,000).
- Dolphin LLM serves one open-weight model without the refusal behaviour of mainstream models.
- Dolphin LLM is prepaid by card or crypto with no subscription; failed requests are free.
Prices side by side
| Model | Input / 1M | Output / 1M | Context |
|---|---|---|---|
| Dolphin LLM (uncensored) | $0.15 | $0.50 | 64,000 |
| OpenAI GPT-6 Luna | $0.10 | $0.50 | 1,050,000 |
| OpenAI GPT-6.1 Terra | $1.00 | $5.00 | 1,050,000 |
For a typical chat turn of 4,000 tokens in and 2,000 out, GPT-6 Luna costs $0.0014, Dolphin LLM $0.0016 and GPT-6.1 Terra $0.014. On price alone, OpenAI's budget model is slightly cheaper on input and equal on output.
Context window
OpenAI's current models accept up to 1,050,000 tokens. Dolphin LLM accepts 64,000 tokens, prompt and answer together. If you feed whole books or very large codebases in one request, OpenAI fits that better. For chat, extraction from long documents and agents with tool calls, 64,000 tokens is usually enough.
The model
OpenAI's models are closed and follow OpenAI's usage policies, including refusals for topics the provider considers sensitive. Dolphin LLM serves one open-weight model tuned to answer lawful adult requests without refusals, which matters for fiction, research and moderation work where refusals break the workflow.
Data
OpenAI states that API data is not used for training unless you opt in, and that abuse-monitoring logs are kept for up to 30 days; zero data retention is available to approved customers. Dolphin LLM does not use prompts or answers for training and keeps usage counters (requests, tokens, cost) per key.
Billing
Dolphin LLM is prepaid: you top up by card or crypto (USDT, USDC), credit never expires, failed requests are free and the balance never goes negative. There is a $0.50 trial credit without a card.
Which one to choose
- Choose OpenAI if you need very long context, images or audio, or the lowest input price.
- Choose Dolphin LLM if your workload hits refusals, if you want an open-weight model behind the same SDK, or if you prefer prepaid billing with crypto.
Switching takes three lines: see the docs.
Sources
Prices change. Check each provider's page before you decide.
Questions and answers
Can I use the OpenAI SDK with Dolphin LLM?
Yes. Set the base URL to the Dolphin LLM endpoint, use your Dolphin LLM key and the model name "uncensored". The rest of your code stays the same.
Is Dolphin LLM cheaper than OpenAI?
Not across the board. GPT-6 Luna is $0.10 per 1M input tokens against $0.15, and both cost $0.50 per 1M output tokens. GPT-6.1 Terra is ten times more expensive than either.