ai.geraikita.com · OpenAI-compatible endpoint

One API for
every top AI model.

Claude, GPT, DeepSeek, Qwen, Kimi, and GLM through a single endpoint. Pay per token with GeraiPoints — no subscription, no expiring quota, genuine models with no tricks.

1 GP = Rp1 · top up GeraiPoints16+ models99.9% automatic failover
example — chat completion
# fully compatible with the OpenAI SDK
curl https://ai.geraikita.com/v1/chat/completions \
  -H "Authorization: Bearer gk-xxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-sonnet-5",
    "messages": [
      {"role":"user","content":"Halo!"}
    ]
  }'

# response
{
  "model": "claude-sonnet-5",
  "usage": { "prompt_tokens": 12, "completion_tokens": 8 }
}
Genuine models from official providers — no impersonation
Claude Opus 5Claude Opus 4.8Claude Opus 4.6Claude Opus 4.7Claude Sonnet 5Claude Sonnet 4.6Claude Haiku 4.5Claude Fable 5DeepSeek V4 ProDeepSeek V4.1 FlashDeepSeek V4 FlashClaude Opus 5 ThinkingClaude Sonnet 5 ThinkingGLM-5.1GLM-5 TurboGLM 5.3GLM 5.2GPT 5.4GPT 5.5GPT 5.5-xhighGPT 5.6 LunaGPT 5.6 SolGPT 5.6 TerraQwen3 MaxQwen3 PlusQwen3 FlashQwen3 Coder PlusKimi K2.7 CodeKimi K3Claude Opus 5 (API Token)Claude Opus 5.5 (API Token)GPT 6 AstraGPT 6 SolGPT 6 LunaClaude Opus 5.5Claude Fable 5.1GPT 6.1 SolGemini 4 Argon (API Token)GLM 5.3 FlashQwen 3.8 FlashQwen 3.8 MaxGemini 4 ArgonClaude Sonnet 5 (API Token)Claude Opus 4.8 (API Token)Claude Sonnet 4.6 (API Token)Claude Opus 4.7 (API Token)Claude Opus 4.6 (API Token)Claude Fable 5 (API Token)Claude Fable 5.1 (API Token)GPT-6 Astra (API Token)GPT-6 Luna (API Token)GPT-6 Sol (API Token)GPT-5.6 Sol (API Token)GPT-5.6 Terra (API Token)GPT-5.6 Luna (API Token)DeepSeek V4 Pro (API Token)DeepSeek V4 Pro 0813 (API Token)DeepSeek V4.1 Flash (API Token)DeepSeek V4 Flash (API Token)GLM 5.3 (API Token)GLM 5.3 Flash (API Token)GLM 5.2 (API Token)DeepSeek V4 Flash 0731 (API Token)
Why Geraikita AI

Built for developers who demand transparency.

Drop-in OpenAI

Just change the base URL. Works out of the box with the OpenAI, Anthropic, and LangChain SDKs.

Metered & honest

Billed per token from real usage. Per-request usage details are in your dashboard.

Genuine models

Qwen is Qwen, Claude is Claude. No cheap models masquerading as something else.

Automatic failover

Provider busy or erroring? Traffic moves to a backup source without you lifting a finger.

Total tokens served
413,060,836,100

See the most active builders & top supporters — names are anonymized for privacy.

View leaderboard

Get started in 2 minutes.

Create an API key, change the base URL, send your first request.