Language Models

All prices are in USD per million tokens and already include any active model discount. Pro and Max plans apply a further discount on top — see Pricing and Usage. The same list is available from the API at GET http://localhost:3000/v1/models and in the browser at Models.

Model ID	Provider	Context	Input / 1M	Output / 1M	Capabilities
`deepseek-v4.1-flash`	DeepSeek	128K	$0.09	$0.36	vision, tools
`deepseek-v4-pro`	DeepSeek	128K	$1.32	$3.96	tools
`deepseek-v4-flash`	DeepSeek	128K	$0.14	$0.56	tools
`hy4-preview`	Tencent	256K	$0.834	$2.501	tools
`hy3`	Tencent	256K	$0.132	$0.528	tools
`mimo-v2.6-pro`	Xiaomi	128K	$0.217	$0.435	vision, tools
`mimo-v2.6-flash`	Xiaomi	128K	$0.014	$0.028	vision, tools
`mimo-v2.5-pro`	Xiaomi	128K	$0.435	$0.87	tools
`glm-5.3-flash`	Zhipu AI	200K	$0.045	$0.15	vision, tools
`glm-5.3-flashx`	Zhipu AI	200K	$0.37	$1.25	vision, tools
`glm-5.3`	Zhipu AI	200K	$1.4	$4.4	tools
`qwen3.8-flash`	Alibaba	1000K	$0.048	$0.141	vision, tools
`qwen3.8-max`	Alibaba	256K	$2	$6	vision, tools
`minimax-m3`	MiniMax	1000K	$0.3	$1.2	vision, tools
`kimi-k3`	Moonshot	256K	$3	$15	vision, tools
`claude-opus-5.5`	Anthropic	200K	$4	$20	vision, tools
`claude-sonnet-5.5`	Anthropic	200K	$2	$10	vision, tools
`gpt-6-sol`	OpenAI	400K	$2.5	$10	vision, tools
`gpt-5.6-luna`	OpenAI	400K	$0.25	$2	vision, tools
`gemini-3.8-pro`	Google	1000K	$2	$12	vision, tools

Choosing a model

Everyday chat and agents — deepseek-v4.1-flash, glm-5.3-flash or qwen3.8-flash: fast and inexpensive.
Coding and complex reasoning — deepseek-v4-pro, kimi-k3 or glm-5.3.
Images — choose a model with vision support (marked on the Models page), or let auto pick one.
Not sure? — use auto.

Model IDs

Use the ID from the first column in the model field. Display names (for example DeepSeek-V4.1-Flash) are also accepted, case-insensitively.

Auto Mode

Pricing and Usage