Choose the model. Keep the interface.
Route workloads across DeepSeek, Qwen, Kimi, GLM, and MiniMax through one OpenAI-compatible API. Change the model field — same endpoint, same SDK, same billing. Every model page includes per-token pricing, context window specs, benchmark data, and a working API example.
Available on BestLLM
LIVEThese 7 models are live and accessible through the OpenAI-compatible API at api.bestllm.dev/v1. Each route includes regional routing, workspace-level spend controls, and per-token billing. Click through for full pricing breakdowns and integration examples.
DeepSeek Chat
Frontier reasoning at a fraction of the cost.
DeepSeek Reasoner
Extended chain-of-thought for complex problems.
Qwen3.8 Max
Alibaba's 2.4T-parameter flagship, now with 1M context.
Qwen Plus
Balanced performance and cost for production workloads.
Kimi K3
Open frontier reasoning with a 1M-token context.
GLM-5
Zhipu's current flagship for coding and agents.
MiniMax M3
Ultra-cheap 1M-context model built for agents.
Model Reference Pages
REFERENCE ONLYPrevious-generation and upcoming models we cover with pricing guides and integration references. These models are not yet available through the BestLLM API — they are editorial content to help you track the model landscape. For production workloads, use the live models above.
Qwen Max (2024)
Previous-generation Alibaba flagship — superseded by Qwen3.8 Max.
Kimi K2 (2025)
Previous-gen long-context MoE — superseded by Kimi K3.
GLM-4 Plus (2024)
Previous-gen Zhipu flagship — superseded by GLM-5.
DeepSeek V3
The open-source model that changed the cost equation.
MiniMax H3
Open-source omni-modal video generation with native audio.
MiniMax H3 Max
Post-trained H3 variant optimized for speed and prompt adherence.
Start Building with BestLLM
One API key. Every supported model. OpenAI-compatible format. Switch models by changing one field.