Models
Choose the right model.
Find an available model, adjust its reasoning effort and understand what a request costs.
Start with your model list
Request /models using the same key as your app. Copy an exact data[].id into the client’s Model ID field. Available IDs can depend on your access; use the returned list before choosing a model from the reference below.
curl --fail-with-body "https://api.lumaos.cloud/v1/models" \
-H "Authorization: Bearer $LUMA_CLOUD_API_KEY"A display name such as “Terra” is not a model ID. Do not add a provider prefix unless your client’s setup guide requires it. Refresh the list if a saved model becomes unavailable.
Reasoning effort
Effort controls how much reasoning the model can use. Higher settings may take longer and consume more tokens. Start with medium and increase it when your task benefits.
| Model family | Base ID | Effort options |
|---|---|---|
| GPT-5.5 | gpt-5.5 | low, medium, high, xhigh |
| GPT-5.6 Sol | gpt-5.6-sol | low, medium, high, xhigh, max, ultra |
| GPT-5.6 Terra | gpt-5.6-terra | low, medium, high, xhigh, max, ultra |
| GPT-5.6 Luna | gpt-5.6-luna | low, medium, high, xhigh, max |
| GPT-6 Astra | gpt-6-astra | low, medium, high, xhigh, max, ultra |
| GPT-6 Sol | gpt-6-sol | low, medium, high, xhigh, max, ultra |
| GPT-6 Luna | gpt-6-luna | low, medium, high, xhigh, max |
This is a reference for model families. Use only IDs and options available to your key. The base IDs use medium when no effort is specified.
Chat Completions uses reasoning_effort; Responses uses reasoning.effort. If you choose an ID that already ends with an effort, keep any separate effort setting identical or omit it. Conflicting settings return an error.
Context and output
The context window is the amount of conversation and input the model can consider. The output limit is the maximum requested response length, including reasoning where applicable. Longer context is not a guarantee that every client can send that much text.
| Model | Context tokens | Maximum output |
|---|---|---|
| GPT-5.5 | 272,000 | 128,000 |
| GPT-5.6 Sol | 372,000 | 128,000 |
| GPT-5.6 Terra | 372,000 | 128,000 |
| GPT-5.6 Luna | 372,000 | 128,000 |
| GPT-6 Astra | 372,000 | 128,000 |
| GPT-6 Sol | 372,000 | 128,000 |
| GPT-6 Luna | 372,000 | 128,000 |
These guides cover text and tool-calling workflows. Tool execution happens in your client and requires its permissions. Do not enable embeddings, autocomplete or media features just because a settings screen offers them.
API Meteor pricing
Our fixed rates are up to 60% below OpenAI API Standard. The off percentage compares fresh-input and output rates with OpenAI Standard for the same model and context. Rates below are in USD per one million tokens. Input is what you send, output is what the model generates, and cached input is eligible reused input. Reasoning can increase token usage without changing the listed per-token rate.
| Model | Input | Cached input | Cache write | Output | vs OpenAI Standard |
|---|---|---|---|---|---|
| GPT-6 Astra | $6 | $0.6 | $7.5 | $30 | 40% off |
| GPT-6 Sol | $1.2 | $0.12 | $1.5 | $6 | 40% off |
| GPT-6 Luna | $0.06 | $0.006 | $0.075 | $0.3 | 40% off |
| GPT-5.6 Sol | $2 | $0.2 | $2.5 | $10 | 50% off |
| GPT-5.6 Terra | $0.8 | $0.08 | $1 | $4.8 | 60% off |
| GPT-5.6 Luna | $0.08 | $0.008 | $0.1 | $0.48 | 60% off |
| GPT-5.5 | $3 | $0.3 | $3 | $18 | 40% off |
For requests above 272,000 input tokens, input, cached-input and cache-write rates are multiplied by 2 and output rates by 1.5 for the full request. A dash means no rate is published for that category. The total for each request is rounded up once to one microdollar ($0.000001). Reasoning is part of provider output and is charged once.
Wallet keys spend your prepaid API balance. Subscription keys and SuperGPT use their applicable subscription allowance instead. Savings in the Wallet activity view compare settled charges with the saved OpenAI Standard rates for the same tokens. Older requests without a saved reference are excluded. OpenAI publishes no GPT-5.5 cache-write price, so requests containing that category are also excluded from the monetary savings comparison. Review both separately in Usage; see billing and additional quota for the difference.