GPT-6 Astra and GPT-6.1 Sol

gpt-6-astra supports POST /v1/responses, streaming and tool calls through the upstream provider subscription route. Eligible tokenized subscriptions use a final 1.25× ordinary input/output multiplier before independent customer discounts; existing cache accounting applies.

gpt-6.1-sol supports POST /v1/responses, streaming and tool calls. GPT-6.1 Sol is a compatibility model powered by MiniMax-M3.1-Flash through upstream provider. It is not native OpenAI GPT-6.1 inference.

GPT-6.1 Sol inherits exactly the effective GPT-6 Sol token billing of your plan and subscription, including the existing upstream provider reduction once. GPT-6 Sol and GPT-6.1 Sol remain distinct model IDs. Both require an eligible subscription; automatic PAYG fallback is unavailable.

Max Unlimited uses its existing separate weighted-request accounting: clamp(ceil((ordinary_input + output + 0.1 × cache_read) / 29000), 1, 50); image requests have a minimum of 5 units and failed requests use zero units. No model multiplier applies. Standard allowances remain 2,000 weighted units/5h and 10,000/7d; existing trial allowances remain as granted.

Endpoint

StartupAPI supports the core Anthropic Messages request, response, and SSE event shapes used by Claude Code and direct API callers. Compatibility is limited to behavior implemented and verified by StartupAPI; optional Anthropic services and beta features can require additional gateway support.

Authentication

StartupAPI requires your StartupAPI API key in Authorization: Bearer. X-Api-Key is not accepted by the StartupAPI authentication middleware. Missing, malformed, inactive, expired, or invalid keys are rejected before model usage is reserved.

Example request

curl https://startupapi.io/v1/messages \
  -H "Authorization: Bearer YOUR_STARTUPAPI_API_KEY" \
  -H "content-type: application/json" \
  -H "anthropic-version: 2023-06-01" \
  -d '{
    "model": "claude-sonnet-4-6",
    "max_tokens": 1024,
    "messages": [
      { "role": "user", "content": "Hello, Claude." }
    ]
  }'

Representative response shape

{
  "id": "msg_...",
  "type": "message",
  "role": "assistant",
  "model": "claude-sonnet-4-6",
  "content": [{ "type": "text", "text": "Hello." }],
  "stop_reason": "end_turn",
  "stop_sequence": null,
  "usage": {
    "input_tokens": 10,
    "cache_creation_input_tokens": 0,
    "cache_read_input_tokens": 0,
    "output_tokens": 4
  }
}

Streaming request

Streaming is enabled in production but is also controlled per API key. New keys currently receive streaming access; an older or restricted key can return a permission error until streaming is enabled for that key.

curl --no-buffer https://startupapi.io/v1/messages \
  -H "Authorization: Bearer YOUR_STARTUPAPI_API_KEY" \
  -H "content-type: application/json" \
  -H "anthropic-version: 2023-06-01" \
  -d '{
    "model": "claude-sonnet-4-6",
    "max_tokens": 1024,
    "stream": true,
    "messages": [{ "role": "user", "content": "Hello, Claude." }]
  }'

Representative event sequence

event: message_start
data: {"type":"message_start","message":{"type":"message","usage":{"input_tokens":10,"output_tokens":0}}}

event: content_block_delta
data: {"type":"content_block_delta","index":0,"delta":{"type":"text_delta","text":"Hello"}}

event: message_stop
data: {"type":"message_stop"}

Common body fields