Thorbase

An LLM API aggregation platform offering plug-and-play access to multiple AI model providers including DeepSeek, OpenAI, Google, Zhipu GLM, Qwen, Moonshot Kimi, and MiniMax, with customizable routing, fallback mechanisms, and per-token pricing.

DeepSeek V3.2 GLM-5 Kimi K2.5

Thorbase offers 12 LLM API models.

API pricing per token ranges from $0.200 to $1.40/M (input).

Speed benchmark average: 43 tok/s.

Avg Speed42.73 tok/s

Latency17.14 s

Total Tests50

Models4

Updated4/17/2026

Created At4/12/2026

Website

API Endpoints

dashboard.thorbase.com

Supported Models

Model	Speed	Latency	Tests
deepseek/deepseek-v3.2	56.13 tok/s	1.98s	5
z-ai/glm-5	42.85 tok/s	24.10s	15
qwen/qwen3.5-397b-a17b	41.34 tok/s	21.79s	15
moonshotai/kimi-k2.5	39.54 tok/s	10.60s	15

Available Models

8 models available (last updated: 4/16/2026)

deepseek/deepseek-v3.1customdeepseek/deepseek-v3.2customminimax/minimax-m2.7custommoonshotai/kimi-k2.5customqwen/qwen3.5-397b-a17bcustomqwen/qwen3.6-pluscustomz-ai/glm-5customz-ai/glm-5.1custom

Thorbase

API Endpoints

Supported Models

Available Models

Thorbase

API Endpoints

Leaderboard Rankings

Supported Models

Available Models

Leaderboard Rankings