🛰️ ORBIT

Free 2B tokens/mo — Claude-only provider, requires User-Agent header

🛰️

ORBIT

Speed: ⚡ 25 tok/s avg
Models: 18
Price: 2B tokens/mo free
Status: ⚠️ Claude Only
Avg Speed
25 tok/s
Models
18
Price
Free 2B/mo
Best For
Free Claude access

💰 Plan & Pricing

✅
Free 2B Tokens/Month
Massive free allocation. Best for free Claude access. Only Claude models work reliably — Gemini models return HTTP 500.

🔑 API Key

sk-orb...e0d3

🌐 Endpoint

https://api.orbit-provider.com/v1/chat/completions

📦 Models (18 total — Claude primary)

ModelSpeedCategoryReliability
claude-opus-4-6⚡ 29 tok/sReasoning✅ Reliable
claude-sonnet-4-5⚡ 25 tok/sChat✅ Reliable
claude-3.5-sonnet⚡ 28 tok/sChat✅ Reliable
claude-haiku-4⚡ 48 tok/sChat✅ Reliable
gpt-4o-mini⚡ 40 tok/sChat✅ Works
deepseek-r1-0528⚡ 28 tok/sReasoning⚠️ May be slow
qwen3-235b⚡ 32 tok/sChat/Coding⚠️ May be slow
llama-4-maverick⚡ 30 tok/sChat⚠️ May be slow
gemma-3-27b⚡ 35 tok/sChat⚠️ May be slow
claude-3-opus⚡ 22 tok/sReasoning✅ Reliable

+ 8 more models listed — all Gemini models return HTTP 500, only Claude models work reliably

💻 cURL Example

curl -X POST https://api.orbit-provider.com/v1/chat/completions \
  -H "Authorization: Bearer sk-orb...e0d3" \
  -H "User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-sonnet-4-5",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

🐍 Python Example

from openai import OpenAI
import httpx

client = OpenAI(
    api_key="sk-orb...e0d3",
    base_url="https://api.orbit-provider.com/v1",
    default_headers={
        "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36"
    }
)

response = client.chat.completions.create(
    model="claude-sonnet-4-5",
    messages=[{"role": "user", "content": "Hello!"}]
)
print(response.choices[0].message.content)

⚠️ Pitfalls & Notes

🚨
MUST Include User-Agent! — Without User-Agent: Mozilla/5.0... header, all requests return 403 Forbidden. This is a hard requirement.
🚨
Gemini Models Fail — All Gemini models return HTTP 500. Only Claude models work reliably.
✅
Free Claude Access — 2 billion tokens/month of Claude for free is the largest free Claude allocation available anywhere.

🏷️ Categories

Chat Coding Reasoning