Chinese AI Models

Open experimental architecture preview behind the production Qwen3.8-Flash line, with 125B parameters, 6B activated, native vision and efficient long-context attention.

Context

262K

Providers

1

Input from

-

Open MIT-licensed multimodal GLM model with 320B total and 18B active parameters, a million-token context, hybrid sparse and linear attention, and first-party API access.

Context

1.05M

Providers

2

Input from

-

The original hosted GLM-5.3 flagship for complex software engineering, long-horizon agents and cybersecurity analysis. Its Coding Plan access and text-only profile are distinct from the newer open, multimodal GLM-5.3-Flash variant.

Context

1M

Providers

1

Input from

-

Apache-2.0 dense 27B multimodal Qwen model for text, image and video understanding with reasoning and tool use.

Context

262K

Providers

1

Input from

-

The hosted V4 Pro snapshot currently served by DeepSeek's official API, with million-token context, 384K output and peak/off-peak billing.

Context

1M

Providers

2

Input from

$0.66

Qwen's first open Max-class text model, a sparse 2.4T-parameter MoE with 95B active parameters for generation, coding and agent work.

Context

262K

Providers

1

Input from

-

Qwen Studio's current flagship model for coding, professional work, multimodal understanding and long-horizon agents.

Context

1M

Providers

2

Input from

-

Multimodal Kimi model with million-token context and max-effort reasoning for long agent work.

Context

1.05M

Providers

2

Input from

$3

Open reasoning model for coding, instruction following and agent tasks.

Context

256K

Providers

2

Input from

$0.13

Open flagship GLM for long-horizon coding agents and million-token context work.

Context

1M

Providers

2

Input from

$0.56

Coding-focused Kimi model designed for long-horizon repository work with controlled reasoning.

Context

262K

Providers

2

Input from

$0.95

Multimodal Qwen workhorse for long-context agents, visual inputs and coding.

Context

1M

Providers

2

Input from

$0.32

Open multimodal model for long-context coding, perception and agent planning.

Context

1M

Providers

2

Input from

$0.23

Open multimodal flash model for fast agents, coding and visual prompts.

Context

256K

Providers

2

Input from

$0.18

Qwen frontier model tuned for agent frameworks, coding assistants and long tasks.

Context

1M

Providers

2

Input from

$1.48

Fast DeepSeek V4 lane for economical reasoning, coding and long-context work.

Context

1M

Providers

2

Input from

$0.0882

Open MoE flagship with million-token context for coding and long agent runs.

Context

1M

Providers

2

Input from

$0.43

Open multimodal MiMo model for coding agents and long-context automation.

Context

1.05M

Providers

2

Input from

$0.14

Open MiMo Pro tier for long-context reasoning and coding-agent execution.

Context

1.05M

Providers

2

Input from

$0.43

Preview generation of Tencent's open reasoning model for coding and agent evaluation.

Context

256K

Providers

2

Input from

$0.063

Open GLM coding model for agentic engineering, terminals and repository generation.

Context

200K

Providers

2

Input from

$0.95

Fast vision model for screenshots, documents and multimodal agent tasks.

Context

200K

Providers

2

Input from

$1.2

Omni model for text, image, video, audio, PDFs and agent workflows.

Context

262K

Providers

1

Input from

$0.4

Open model for coding agents, office automation and complex tool environments.

Context

205K

Providers

2

Input from

$0.25

Open flash model for quick reasoning, coding assistance and efficient deployment.

Context

256K

Providers

2

Input from

$0.1

Hosted Qwen coding model for software agents, repository edits and long-context code work.

Context

1.05M

Providers

2

Input from

$0.65