Skip to main content
POST
/
v1
/
inference
/
completions
curl -X POST https://api.gravixlayer.com/v1/inference/completions \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer $GRAVIXLAYER_API_KEY' \
  -d '{
    "model": "meta-llama/llama-3.1-8b-instruct",
    "prompt": "Hello! Tell me about AI."
  }'
{
  "data": {
    "id": "<string>",
    "object": "text.completion",
    "created": 123,
    "model": "<string>",
    "choices": [
      {
        "text": "<string>",
        "index": 123,
        "logprobs": {},
        "finish_reason": "<string>"
      }
    ]
  }
}

Authorizations

​
Authorization
string
header
required

API key authentication. Get your API key from the Gravix Layer Dashboard.

Body

application/json
​
model
string
required

Model identifier

Example:

"meta-llama/llama-3.1-8b-instruct"

​
prompt
string
required

Prompt to complete

​
max_tokens
integer
default:16

Maximum tokens to generate

​
temperature
number
default:1

Sampling temperature

Required range: 0 <= x <= 2
​
top_p
number
default:1

Nucleus sampling

Required range: 0 <= x <= 1
​
n
integer
default:1

Number of completions

​
stream
boolean
default:false

Whether to stream the response

​
logprobs
integer | null

Include top N log probabilities

​
echo
boolean
default:false

Echo the prompt in response

​
stop

Stop sequences

​
presence_penalty
number
default:0

Presence penalty

Required range: -2 <= x <= 2
​
frequency_penalty
number
default:0

Frequency penalty

Required range: -2 <= x <= 2
​
best_of
integer
default:1

Generate N completions, return best

​
logit_bias
object

Modify token probabilities

​
user
string | null

User identifier

Response

Completion response

​
data
object