API Reference

Base URL: https://api.featherless.ai/v1

POST /v1/chat/completions

Chat and text generation with conversation history support.

Parameters

Response

Examples

curl https://api.featherless.ai/v1/chat/completions \
  -H "Authorization: Bearer sk-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Qwen/Qwen2.5-72B-Instruct",
    "messages": [
      {"role": "system", "content": "You are a helpful assistant."},
      {"role": "user", "content": "Explain quantum computing in one sentence."}
    ],
    "temperature": 0.7,
    "max_tokens": 200
  }'

<?php
$ch = curl_init('https://api.featherless.ai/v1/chat/completions');
curl_setopt_array($ch, [
    CURLOPT_POST => true,
    CURLOPT_RETURNTRANSFER => true,
    CURLOPT_HTTPHEADER => [
        'Authorization: Bearer sk-YOUR_KEY',
        'Content-Type: application/json'
    ],
    CURLOPT_POSTFIELDS => json_encode([
        'model' => 'Qwen/Qwen2.5-72B-Instruct',
        'messages' => [
            ['role' => 'system', 'content' => 'You are a helpful assistant.'],
            ['role' => 'user', 'content' => 'Explain quantum computing in one sentence.']
        ],
        'temperature' => 0.7,
        'max_tokens' => 200
    ])
]);
$response = json_decode(curl_exec($ch), true);
echo $response['choices'][0]['message']['content'];

const response = await fetch('https://api.featherless.ai/v1/chat/completions', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer sk-YOUR_KEY',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    model: 'Qwen/Qwen2.5-72B-Instruct',
    messages: [
      { role: 'system', content: 'You are a helpful assistant.' },
      { role: 'user', content: 'Explain quantum computing in one sentence.' }
    ],
    temperature: 0.7,
    max_tokens: 200
  })
});
const data = await response.json();
console.log(data.choices[0].message.content);

Best for: chat, coding assistance, reasoning, OCR (with vision models)

POST /v1/completions

Raw text completion (no chat formatting). Uses prompt (string) instead of messages. Same parameters otherwise.

Key Difference from Chat

Response

Example

curl https://api.featherless.ai/v1/completions \
  -H "Authorization: Bearer sk-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Qwen/Qwen2.5-72B-Instruct",
    "prompt": "The meaning of life is",
    "max_tokens": 50,
    "temperature": 0.8
  }'

POST /v1/tokenize

Count tokens for a given input without generating a response.

Request Body

Example

curl https://api.featherless.ai/v1/tokenize \
  -H "Authorization: Bearer sk-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Qwen/Qwen2.5-72B-Instruct",
    "messages": [{"role": "user", "content": "Hello, how are you?"}]
  }'

GET /v1/models

List all available text models.

Query Parameters

Example

curl https://api.featherless.ai/v1/models \
  -H "Authorization: Bearer sk-YOUR_KEY"

curl https://api.featherless.ai/v1/models?available_on_current_plan=1 \
  -H "Authorization: Bearer sk-YOUR_KEY"

Vision models accept images alongside text in the messages array. Images can be provided via URL or base64.

Supported Vision Models

Sending an Image

Use the content array format in a message. Each content item has a type of text or image_url.

Via URL

{
  "model": "google/gemma-3-27b-it",
  "messages": [{
    "role": "user",
    "content": [
      { "type": "text", "text": "What is in this image?" },
      { "type": "image_url", "image_url": { "url": "https://example.com/photo.jpg" } }
    ]
  }]
}

Via Base64

{
  "model": "google/gemma-3-27b-it",
  "messages": [{
    "role": "user",
    "content": [
      { "type": "text", "text": "Extract all text from this image." },
      { "type": "image_url", "image_url": { "url": "data:image/png;base64,iVBORw0KGgo..." } }
    ]
  }]
}

Full cURL Example (OCR)

curl https://api.featherless.ai/v1/chat/completions \
  -H "Authorization: Bearer sk-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "google/gemma-3-27b-it",
    "messages": [{
      "role": "user",
      "content": [
        { "type": "text", "text": "Extract all text from this image. Output only the text, nothing else." },
        { "type": "image_url", "image_url": { "url": "https://example.com/receipt.jpg" } }
      ]
    }],
    "max_tokens": 500
  }'

Base URL: https://api.arliai.com/v1

Arli AI text endpoints are OpenAI-compatible with additional sampling and guidance parameters. Offers 117 text models.

POST /v1/chat/completions

Chat and text generation. Same base parameters as Featherless, plus Arli-specific extras:

Standard Parameters

Arli-Specific Parameters

Example

curl https://api.arliai.com/v1/chat/completions \
  -H "Authorization: Bearer sk-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Qwen/Qwen2.5-72B-Instruct",
    "messages": [
      {"role": "system", "content": "You are a coding expert."},
      {"role": "user", "content": "Write a Python function to sort a list."}
    ],
    "temperature": 0.6,
    "top_p": 0.9,
    "max_tokens": 500,
    "repetition_penalty": 1.1
  }'

POST /v1/completions

Raw text completion. Uses prompt instead of messages. Same extended parameters as chat.

curl https://api.arliai.com/v1/completions \
  -H "Authorization: Bearer sk-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Qwen/Qwen2.5-72B-Instruct",
    "prompt": "Complete this: The future of AI is",
    "max_tokens": 100,
    "temperature": 0.8
  }'

GET /v1/models

List all 117 available text models.

curl https://api.arliai.com/v1/models \
  -H "Authorization: Bearer sk-YOUR_KEY"

POST /v1/tokenize

Count tokens for a given input.

curl https://api.arliai.com/v1/tokenize \
  -H "Authorization: Bearer sk-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Qwen/Qwen2.5-72B-Instruct",
    "messages": [{"role": "user", "content": "Count my tokens"}]
  }'

Base URL: https://api.arliai.com/v1

Arli AI provides Stable Diffusion-based image generation with 74 image models, including FLUX, SDXL, SD 1.5, and anime/pixel art variants.

POST /v1/txt2img

Generate images from text prompts.

Parameters

Response

Example

curl https://api.arliai.com/v1/txt2img \
  -H "Authorization: Bearer sk-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "sd_model_checkpoint": "FLUX.2-klein-4B",
    "prompt": "a serene mountain lake at sunset, photorealistic, 4K",
    "negative_prompt": "blurry, low quality, distorted",
    "steps": 25,
    "sampler_name": "euler",
    "width": 1024,
    "height": 768,
    "seed": -1,
    "cfg_scale": 7
  }' | python3 -c "
import sys, json, base64
data = json.load(sys.stdin)
img = base64.b64decode(data['images'][0])
with open('output.png', 'wb') as f: f.write(img)
print('Saved output.png')
"

POST /v1/img2img

Transform an existing image based on a text prompt.

Parameters

Example

curl https://api.arliai.com/v1/img2img \
  -H "Authorization: Bearer sk-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "sd_model_checkpoint": "FLUX.2-klein-4B",
    "prompt": "turn daytime scene into night with stars",
    "init_images": ["'"$(base64 -w0 input.png)"'"],
    "denoising_strength": 0.6,
    "steps": 25,
    "cfg_scale": 7
  }'

POST /v1/upscale-img

Upscale an image using ESRGAN models.

Parameters

Response

Example

curl https://api.arliai.com/v1/upscale-img \
  -H "Authorization: Bearer sk-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "image": "'"$(base64 -w0 photo.png)"'",
    "upscaler_1": "ESRGAN_4x"
  }' | python3 -c "
import sys, json, base64
data = json.load(sys.stdin)
img = base64.b64decode(data['images'][0])
with open('upscaled.png', 'wb') as f: f.write(img)
print('Saved upscaled.png (4x)')
"

GET /sdapi/v1/sd-models

List all 74 available image generation models.

curl https://api.arliai.com/sdapi/v1/sd-models \
  -H "Authorization: Bearer sk-YOUR_KEY"

GET /v1/img-samplers

List available sampler algorithms.

curl https://api.arliai.com/v1/img-samplers \
  -H "Authorization: Bearer sk-YOUR_KEY"

GET /v1/upscalers

List available upscaler models.

curl https://api.arliai.com/v1/upscalers \
  -H "Authorization: Bearer sk-YOUR_KEY"

GET /v1/img-options

Get current image generation options and defaults.

curl https://api.arliai.com/v1/img-options \
  -H "Authorization: Bearer sk-YOUR_KEY"

GET /v1/parallel-requests

Check rate limits for image generation.

curl https://api.arliai.com/v1/parallel-requests \
  -H "Authorization: Bearer sk-YOUR_KEY"

Featherless AI Vision

Images accepted via URL or base64 (data URI format).

Arli AI Vision

Images accepted via base64 ONLY (URL input is not supported).

OCR Best Practice

Use google/gemma-3-27b-it on Featherless with a clear OCR prompt:

{
  "model": "google/gemma-3-27b-it",
  "messages": [{
    "role": "user",
    "content": [
      { "type": "text", "text": "Extract all text from this image. Preserve formatting and line breaks." },
      { "type": "image_url", "image_url": { "url": "https://example.com/document.png" } }
    ]
  }],
  "max_tokens": 2000,
  "temperature": 0.1
}

Pick your goal and find the right endpoint and example.

Common Error Codes

Error Response Format

Rate Limiting Behavior

• Featherless: Subject to per-plan request limits. Returns 429 with Retry-After header.
• Arli AI Text: Standard rate limits apply per API key.
• Arli AI Image: Maximum 6 concurrent requests. Check GET /v1/parallel-requests for remaining capacity.

Best Practices

• Always handle 429 by reading the Retry-After header and waiting before retrying
• Use exponential backoff for 500 errors (wait 1s, 2s, 4s...)
• Check GET /v1/models to verify model availability before sending requests
• Use POST /v1/tokenize to count tokens before sending large prompts

Interactive web UIs for testing all providers without writing code.