DeepSeek V3.2
DeepSeek
GPT-4-class general model. The most cost-effective model in the world for code & chat.
- Context
- 64K
- Category
- Chat
- Input
- $0.14/M
- Output
- $0.28/M
ToolsJSON modeStreamingFunctions
0% markup Β· pass-through
Models
Every frontier model. One endpoint.
Same OpenAI SDK. Same `model` field. Smart router picks the lowest-cost healthy channel. Below: every model on the network, with the actual rate, context, and capabilities. Click any card to copy its model id.
Capability
Origin
Showing 15 models
CN
DeepSeek R1
DeepSeek
Open reasoning model rivalling o1 on math & logic, at 1/30 the price.
- Context
- 64K
- Category
- Reasoning
- Input
- $0.55/M
- Output
- $2.19/M
Streaming
0% markup Β· pass-through
Kimi K2.5
Moonshot
Long-context tool-use champion. Strong agentic behaviour, free 128K context.
- Context
- 128K
- Category
- Chat
- Input
- $0.6/M
- Output
- $2.5/M
ToolsJSON modeStreamingFunctions
0% markup Β· pass-through
GLM-4.6
Zhipu
ModelDescriptions.glm-4.6
- Context
- 128K
- Category
- Code
- Input
- $0.5/M
- Output
- $1.5/M
ToolsJSON modeStreaming
0% markup Β· pass-through
Qwen3 Max
Alibaba
256K context. Strong on multilingual and bilingual reasoning.
- Context
- 256K
- Category
- Chat
- Input
- $0.6/M
- Output
- $2.4/M
ToolsStreaming
0% markup Β· pass-through
Qwen3 Coder
Alibaba
Coder-specialised Qwen variant. SOTA on HumanEval-CN.
- Context
- 128K
- Category
- Code
- Input
- $0.3/M
- Output
- $1.5/M
ToolsStreaming
0% markup Β· pass-through
Doubao 1.6 Pro
ByteDance
ModelDescriptions.doubao-1.6-pro
- Context
- 64K
- Category
- Vision
- Input
- $0.4/M
- Output
- $1.2/M
VisionStreamingTools
0% markup Β· pass-through
MiniMax M2.5
MiniMax
ModelDescriptions.minimax-m2.5
- Context
- 200K
- Category
- Chat
- Input
- $0.3/M
- Output
- $1.2/M
Streaming
0% markup Β· pass-through
Claude Opus 4.7
Anthropic
Frontier reasoning. Pass-through pricing β exactly the official rate.
- Context
- 200K
- Category
- Reasoning
- Input
- $15/M
- Output
- $75/M
ToolsJSON modeStreamingPrompt cacheFunctions
0% markup Β· pass-through
Claude Sonnet 4.6
Anthropic
Workhorse Anthropic model. 5Γ cheaper than Opus, ~90% the quality.
- Context
- 200K
- Category
- Chat
- Input
- $3/M
- Output
- $15/M
ToolsJSON modeStreamingPrompt cacheFunctions
0% markup Β· pass-through
GPT-4o
OpenAI
Multimodal frontier. Pass-through pricing.
- Context
- 128K
- Category
- Vision
- Input
- $2.5/M
- Output
- $10/M
VisionToolsJSON modeStreamingFunctions
0% markup Β· pass-through
GPT-5.5
OpenAI
ModelDescriptions.gpt-5.5
- Context
- 256K
- Category
- Chat
- Input
- $5/M
- Output
- $20/M
VisionToolsJSON modeStreamingFunctions
0% markup Β· pass-through
o1
OpenAI
OpenAI flagship reasoning. Pass-through pricing.
- Context
- 200K
- Category
- Reasoning
- Input
- $15/M
- Output
- $60/M
Streaming
0% markup Β· pass-through
Gemini 3 Pro
2M token context. Pass-through pricing.
- Context
- 2M
- Category
- Chat
- Input
- $1.25/M
- Output
- $5/M
VisionToolsJSON modeStreamingFunctions
0% markup Β· pass-through
Kling 1.6
Kling
ModelDescriptions.kling-1.6
- Category
- Video
- Price
- $0.5 per second of video
0% markup Β· pass-through
Don't see the model you need?
We typically add new Chinese models within 7 days of upstream release. Suggest a model on GitHub and we'll prioritise it.