nexusflow
Online

Model Overview

NexusFlow integrates industry-leading large language models, providing OpenAI, Anthropic Messages, Gemini-compatible, Embeddings, Image Generations, and Tasks public interfaces based on model capabilities. Choose the model best suited for your needs.

Protocol Overview

Model Series

Qwen (Tongyi) Series

Alibaba Cloud

Alibaba Cloud's self-developed large models, currently showcasing Qwen3.6 and Qwen3.5 series with outstanding Chinese capability and ultra-long context support

Qwen3.7 MaxQwen3.6 Max PreviewQwen3.6 PlusQwen3.5 PlusQwen3.5 Flash

Claude Series

Differentiated
Anthropic

Anthropic flagship models with million-level context, outstanding reasoning and coding capability. Called via /v1/messages compatible interface

Claude Opus 4.7Claude Sonnet 4.6Claude Haiku 4.5

DeepSeek Series

High Cost-Efficiency
DeepSeek

High-performance reasoning and general models, outstanding coding capability, suited for complex tasks and high-concurrency scenarios

DeepSeek V4 ProDeepSeek V4 FlashDeepSeek R1DeepSeek V3.2

GLM (Zhipu) Series

Zhipu AI

Domestically leading large language models, comprehensive general capabilities, long context support

GLM 5.1GLM 5GLM 4.7

Kimi Series

Moonshot

Moonshot's large models, strong long-text understanding and reasoning capability

Kimi K2.6Kimi K2.5

MiniMax Series

MiniMax

MiniMax large models, suited for general chat and content creation

MiniMax M2.5MiniMax M2.1

HappyHorse Topic

Latest
Alibaba

VBench-ranked #1 video generation model, supports Text-to-Video, Image-to-Video, already integrated via NexusFlow unified interface

happyhorse-1.0-t2vhappyhorse-1.0-i2vhappyhorse-1.0-r2v

PixVerse Video Models

PixVerse

Professional video generation models, supporting Text-to-Video, Image-to-Video, First & Last Frame, and Reference-to-Video capabilities

PixVerse V6PixVerse V4.5PixVerse V4

Price Comparison

Price unit: per million tokens (CNY)

ModelContextInput PriceOutput PricePositioning
qwen3.7-max1M¥12¥36Flagship
qwen3.6-max-preview262K¥9¥54Flagship
qwen3.6-plus1M¥2¥12Balanced
qwen3.5-plus1M¥0.8¥4.8Balanced
qwen3.5-flash1M¥0.2¥2Ultra-fast
claude-opus-4-71M≈¥34≈¥170Flagship
claude-sonnet-4-61M≈¥20.4≈¥102Balanced
claude-haiku-4-5200K≈¥6.8≈¥34High-Speed
deepseek-v4-pro1M¥12¥24Reasoning Flagship
deepseek-v4-flash1M¥1¥2High-Speed
deepseek-r164K¥4¥16Reasoning
deepseek-v3.2131K¥2¥3General
glm-5.1131K¥5¥20Flagship
glm-5131K¥4¥18Balanced
kimi-k2.6131K¥5¥27Reasoning
kimi-k2.5131K¥4¥21Balanced
MiniMax-M2.51M¥1¥8Balanced

Model Selection Guide

Complex Reasoning and Coding
Qwen3 Max
Flagship capabilities, more balanced, suited for complex reasoning, code generation, and system-level tasks
Daily Chat and Content Creation
Qwen3.5 Plus
Balanced performance and cost, suited for most online chat and business scenarios
Chinese Content Processing
Qwen3.5 Series
Outstanding Chinese understanding and generation capability, supports ultra-long text
High Cost-Efficiency Needs
DeepSeek V3
Open-source model, affordable pricing, outstanding coding capability

Not sure which model to choose?

Try each model for free in Playground and find the solution best suited for you.

Open Playground