AI models

Compare the AI models available on Layest and find the right model for chat or agents.

Providers: 9
EU hosted: 14

18 models

Credit usage per 1M tokens

Model	Hosting	Features				Cache
Amazon
Amazon Nova 2 LiteFast, cost-efficient multimodal model Available in:Assistant	EU	Vision	300K	34.32	287.76	8.58
Amazon Nova ProMultimodal model for analysis Available in:Assistant	EU	Vision	300K	84	336	21
Anthropic
Claude Haiku 4.5Fast, cost-efficient responses Available in:Assistant	EU	VisionReasoning	200K	88	440	8.8
Claude Haiku 5.5Fast, cost-efficient responses Available in:AssistantAgents	EU	VisionReasoning	1M	8.8Tiered	44	0.88
Claude Opus 5.5Advanced reasoning for demanding tasks Available in:Assistant	EU	VisionReasoning	1M	352	1,760	17.6
Claude Sonnet 4.6Balanced reasoning and speed Available in:AssistantAgents	EU	VisionReasoning	1M	264	1,320	26.4
Claude Sonnet 5Balanced model for complex tasks Available in:Assistant	EU	VisionReasoning	1M	176	880	17.6
Claude Sonnet 5.5Balanced model for complex tasks Available in:AssistantAgents	EU	VisionReasoning	1M	176	880	8.8
DeepSeek
DeepSeek V3.2Strong reasoning at a low price Available in:AssistantAgents	EU	Reasoning	164K	59.2	177.6	Not available
MiniMax
MiniMax M2.5Cost-efficient text-only model for agent workflows Available in:AssistantAgents	EU	Reasoning	200K	28.8	115.2	Not available
Mistral AI
Devstral 2Tuned for code and structured output Available in:AssistantAgents	EU	Reasoning	262K	38.4	192	Not available
Moonshot AI
Kimi K2.5Long context with image understanding Available in:AssistantAgents	EU	VisionReasoning	262K	57.6	288	Not available
Kimi K3Long context with image understanding Available in:AssistantAgents	Global	VisionReasoning	1M	240	1,200	24
OpenAI
GPT-6 AstraReasoning model for demanding tasks Available in:AssistantAgents	Global	VisionReasoning	1.05M	800Tiered	4,000	80
GPT-6 LunaFast model for everyday tasks Available in:AssistantAgents	Global	VisionReasoning	1.05M	8Tiered	40	0.8
GPT-6.1 SolGeneral-purpose reasoning model Available in:AssistantAgents	Global	VisionReasoning	1M	160Tiered	800	8
Qwen
Qwen3 235BCheapest capable model for long documents Available in:AssistantAgents	EU	Reasoning	262K	17.6	70.4	Not available
Z.ai
GLM-5Deep reasoning on open weights Available in:AssistantAgents	EU	Reasoning	203K	96	307.2	Not available

Credit usage per 1M tokens

Amazon

Amazon Nova 2 Lite

Fast, cost-efficient multimodal model

Vision

Available in:Assistant

Input: 34.32
Output: 287.76
Context: 300K tokens

Credit usage per 1M tokens

Amazon Nova Pro

Multimodal model for analysis

Vision

Available in:Assistant

Input: 84
Output: 336
Context: 300K tokens

Credit usage per 1M tokens

Anthropic

Claude Haiku 4.5

Fast, cost-efficient responses

VisionReasoning

Available in:Assistant

Input: 88
Output: 440
Context: 200K tokens

Credit usage per 1M tokens

Claude Haiku 5.5

Fast, cost-efficient responses

VisionReasoning

Available in:AssistantAgents

Input: 8.8
Output: 44
Context: 1M tokens

Credit usage per 1M tokens

Claude Opus 5.5

Advanced reasoning for demanding tasks

VisionReasoning

Available in:Assistant

Input: 352
Output: 1,760
Context: 1M tokens

Credit usage per 1M tokens

Claude Sonnet 4.6

Balanced reasoning and speed

VisionReasoning

Available in:AssistantAgents

Input: 264
Output: 1,320
Context: 1M tokens

Credit usage per 1M tokens

Claude Sonnet 5

Balanced model for complex tasks

VisionReasoning

Available in:Assistant

Input: 176
Output: 880
Context: 1M tokens

Credit usage per 1M tokens

Claude Sonnet 5.5

Balanced model for complex tasks

VisionReasoning

Available in:AssistantAgents

Input: 176
Output: 880
Context: 1M tokens

Credit usage per 1M tokens

DeepSeek

DeepSeek V3.2

Strong reasoning at a low price

Reasoning

Available in:AssistantAgents

Input: 59.2
Output: 177.6
Context: 164K tokens

Credit usage per 1M tokens

MiniMax

MiniMax M2.5

Cost-efficient text-only model for agent workflows

Reasoning

Available in:AssistantAgents

Input: 28.8
Output: 115.2
Context: 200K tokens

Credit usage per 1M tokens

Mistral AI

Devstral 2

Tuned for code and structured output

Reasoning

Available in:AssistantAgents

Input: 38.4
Output: 192
Context: 262K tokens

Credit usage per 1M tokens

Moonshot AI

Kimi K2.5

Long context with image understanding

VisionReasoning

Available in:AssistantAgents

Input: 57.6
Output: 288
Context: 262K tokens

Credit usage per 1M tokens

Kimi K3

Global

Long context with image understanding

VisionReasoning

Available in:AssistantAgents

Input: 240
Output: 1,200
Context: 1M tokens

Credit usage per 1M tokens

OpenAI

GPT-6 Astra

Global

Reasoning model for demanding tasks

VisionReasoning

Available in:AssistantAgents

Input: 800
Output: 4,000
Context: 1.05M tokens

Credit usage per 1M tokens

GPT-6 Luna

Global

Fast model for everyday tasks

VisionReasoning

Available in:AssistantAgents

Input: 8
Output: 40
Context: 1.05M tokens

Credit usage per 1M tokens

GPT-6.1 Sol

Global

General-purpose reasoning model

VisionReasoning

Available in:AssistantAgents

Input: 160
Output: 800
Context: 1M tokens

Credit usage per 1M tokens

Qwen

Qwen3 235B

Cheapest capable model for long documents

Reasoning

Available in:AssistantAgents

Input: 17.6
Output: 70.4
Context: 262K tokens

Credit usage per 1M tokens

Z.ai

GLM-5

Deep reasoning on open weights

Reasoning

Available in:AssistantAgents

Input: 96
Output: 307.2
Context: 203K tokens

Credit usage per 1M tokens

Input and output tokens are billed separately. Hosting: The location where the model is executed (inference location). Context: Maximum number of tokens per request. Reasoning: Thinks step-by-step before providing an answer. Image understanding: Capable of interpreting images and documents.

Curious about Layest? Ask your favorite AI