Available Models

List Models

GET https://api.hanzo.ai/v1/models

Returns all available Zen models.

The latest generation. Flagship, reasoning, and code models.

Model	Context	Architecture	Input $/1M	Output $/1M
zen4	202K	~400B Dense	$3.00	$9.60
zen4-ultra	202K	~400B Dense + CoT	$3.00	$9.60
zen4-pro	131K	80B (3B active) MoE	$2.70	$2.70
zen4-max	256K	1.04T (32B active) MoE	$3.60	$3.60
zen4-mini	40K	8B Dense	$0.60	$0.60
zen4-thinking	131K	80B (3B active) MoE + CoT	$2.70	$2.70
zen4-coder	262K	480B (35B active) MoE	$3.60	$3.60
zen4-coder-pro	262K	480B Dense BF16	$4.50	$4.50
zen4-coder-flash	262K	Dense	$1.50	$1.50

Multimodal, vision, safety, and embedding models.

Model	Context	Architecture	Input $/1M	Output $/1M
zen3-omni	202K	~200B Dense Multimodal	$1.80	$6.60
zen3-vl	131K	30B (3B active) MoE VL	$0.45	$1.80
zen3-nano	40K	4B Dense	$0.30	$0.30
zen3-guard	40K	4B Dense	$0.30	$0.30
zen3-embedding	8K	Embedding (3072 dim)	$0.39	--

Task	Recommended Model
General chat	zen4
Maximum reasoning	zen4-ultra
Deep reasoning (CoT)	zen4-thinking
Code generation	zen4-coder
Fast code iteration	zen4-coder-flash
Premium code accuracy	zen4-coder-pro
Image understanding	zen3-vl
Multimodal (text+vision+audio)	zen3-omni
Content moderation	zen3-guard
Text embeddings	zen3-embedding
Edge / mobile	zen3-nano
Budget-friendly	zen4-mini
Extended context docs	zen4-max
High capability	zen4-pro

Budget	Recommended
Free tier ($5)	zen4-mini, zen3-nano
Low cost	zen4-mini, zen3-vl, zen4-coder-flash
Standard	zen4-pro, zen4-coder, zen4-thinking
Premium	zen4, zen4-ultra, zen4-coder-pro

All Zen models are also available as open weights for self-hosting:

Cloud API via Hanzo gives you managed infrastructure, usage tracking, and pay-per-token billing without running your own GPUs.