Best Large Language Models in China
View:
Compare the Top Large Language Models in China as of September 2026
Sort By:
What are Large Language Models in China?
Large language models are artificial neural networks used to process and understand natural language. Commonly trained on large datasets, they can be used for a variety of tasks such as text generation, text classification, question answering, and machine translation. Over time, these models have continued to improve, allowing for better accuracy and greater performance on a variety of tasks. Compare and read user reviews of the best Large Language Models in China currently available using the table below. This list is updated regularly.
-
1
Large Language Models (LLMs) in Gemini Enterprise Agent Platform enable businesses to perform complex natural language processing tasks such as text generation, summarization, and sentiment analysis. These models, powered by massive datasets and cutting-edge techniques, can understand context and generate human-like responses. Gemini Enterprise Agent Platform offers scalable solutions for training, fine-tuning, and deploying LLMs to meet business needs. New customers receive $300 in free credits, allowing them to explore the potential of LLMs in their applications. With these models, businesses can enhance their AI-driven text-based services and improve customer interactions.Starting Price: Free ($300 in free credits)
-
2
Google AI Studio
Google
Google AI Studio provides access to large language models (LLMs) that are capable of understanding and generating human-like text. These models are trained on vast amounts of data and are designed to perform a wide range of language tasks, from translation and summarization to question answering and content generation. By leveraging LLMs, businesses can create applications that understand complex language inputs and produce contextually relevant responses. Google AI Studio also allows users to fine-tune these models, making them highly adaptable to specific use cases or industry requirements.Starting Price: Free -
3
LM-Kit.NET
LM-Kit
LM-Kit.NET lets C# and VB.NET developers integrate large and small language models for natural language understanding, text generation, multi-turn dialogue, and low-latency on-device inference, while its vision language models add image analysis and captioning, its embedding models turn text into vectors for fast semantic search, and its LM-Lit catalog lists every state-of-the-art model with continuous updates, all in one efficient toolkit that stays inside your codebase without revealing any AI origin to the user.Starting Price: Free (Community) or $1000/year -
4
GPT-6 Astra
OpenAI
GPT-6 Astra is OpenAI’s frontier AI model for computer use, software engineering, scientific research, cybersecurity, browsing, and complex professional work. It combines advanced reasoning with agentic capabilities that allow it to navigate software, use tools, conduct research, manipulate data, troubleshoot systems, and complete multistep workflows. Astra is also designed to produce polished documents, spreadsheets, presentations, websites, applications, and other business or technical artifacts while following existing templates and organizational standards. In Codex, the model introduces improved long-running context management that can preserve notes and retrieve information from earlier context windows during extended software engineering tasks. OpenAI positions Astra as its most aligned model to date, with improvements in respecting task boundaries, interpreting user intent, communicating limitations, and avoiding unauthorized actions.Starting Price: $10 per 1M tokens (input) -
5
Claude Fable 5.1
Anthropic
Claude Fable 5.1 is Anthropic’s advanced AI model for coding, knowledge work, research, and long-running agentic tasks. It is designed to improve on Claude Fable 5 with stronger performance across software engineering, scientific research, multidisciplinary reasoning, computer use, business workflows, and complex problem solving. The model can handle extended multi-step work, verify its own results, diagnose difficult software issues, and operate effectively across tool-heavy workflows. Anthropic also reduced cache-read pricing for Fable 5.1, lowering typical usage costs compared with Fable 5 and creating larger savings for highly agentic workloads. Fable 5.1 includes updated safeguards intended to reduce false positives while allowing more legitimate cybersecurity tasks such as vulnerability discovery for defensive purposes. The model is available through Claude products, the Claude API, Amazon Web Services, Google Cloud, and Microsoft Azure.Starting Price: $10 per 1M tokens (input) -
6
GPT-5.6 Sol
OpenAI
GPT-5.6 Sol is a next-generation OpenAI model designed for advanced reasoning, coding, agentic workflows, biology analysis, cybersecurity support, and complex knowledge work. It is part of the GPT-5.6 model family alongside Terra and Luna, with Sol positioned as the flagship model for the most demanding tasks. The model introduces a new max reasoning effort for deeper thinking and an ultra mode that uses subagents to accelerate complex work beyond a single-agent approach. GPT-5.6 Sol shows strong performance in command-line coding workflows, long-horizon security tasks, genomics analysis, vulnerability research, debugging, patch development, and defensive testing. OpenAI pairs the model’s stronger capabilities with layered safeguards, real-time misuse classifiers, account-level review, automated red-teaming, and enterprise controls for sensitive workflows. GPT-5.6 Sol helps developers, enterprises, researchers, and security teams complete sophisticated technical work.Starting Price: $4 per 1M tokens (input) -
7
Claude Mythos 5.1
Anthropic
Claude Mythos 5.1 is Anthropic’s newest Mythos-class model, designed for advanced cybersecurity, biology, scientific research, coding, and long-running knowledge work. It is the same underlying model as Claude Fable 5.1 but uses different safeguards: Fable 5.1 is generally available, while Mythos 5.1 is restricted to trusted access programs with safeguards specifically designed for cybersecurity and life sciences research. The model sets a new performance frontier for agentic coding and demonstrates the strongest cyber capabilities of any Anthropic model released to date. In scientific research, Mythos 5.1 can work with specialized tools and complex workflows across molecular design, computational biology, and other technical domains. In Anthropic’s experiments, it designed high-affinity protein binders across multiple targets and achieved its strongest measured hit rate to date. It also optimized seven open-source protein and genomics deep learning models. -
8
Grok 4.6
SpaceXAI
Grok 4.6 is an xAI model designed for long-running agents, ambitious interactive projects, visual work, coding, research, and knowledge workflows. The model builds on Grok 4.5 with stronger support for multi-step tasks that require sustained reasoning across codebases, information analysis, application development, and work artifact creation. Grok 4.6 can help turn broad product ideas into working first versions by researching domains, structuring applications, implementing core interactions, and refining results through feedback. It is trained across agentic tasks such as knowledge work, general coding, kernel optimization, web development, computer-aided design, and other technical environments. The model is available in Cursor, Grok Build, the xAI API, and partners such as OpenRouter, Vercel, and Cloudflare. Built for developers, builders, and teams working on complex projects, Grok 4.6 helps accelerate coding, agentic workflows, visual applications, and technical execution.Starting Price: $2 per 1M tokens (input) -
9
Claude Opus 5
Anthropic
Claude Opus 5 is Anthropic’s advanced everyday AI model built for coding, knowledge work, problem-solving, visual outputs, and production AI workflows. The model delivers stronger performance than Opus 4.8 at the same base price and is positioned as a cost-effective alternative close to Claude Fable 5 frontier intelligence. Claude Opus 5 supports configurable effort settings so users can optimize for intelligence, speed, or token efficiency. It performs especially well on software engineering, automation, computer use, scientific research, and knowledge work evaluations. The model is available on Claude Max, Claude Pro, Claude API, Claude Code, and other Claude platforms, with Fast mode available at a higher price. Built for developers, researchers, enterprises, and everyday Claude users, Claude Opus 5 helps teams complete complex tasks with stronger verification, careful iteration, and practical cost efficiency.Starting Price: $5 per 1M tokens (input) -
10
Claude Mythos 5
Anthropic
Claude Mythos 5 is Anthropic’s most advanced restricted-access AI model, designed for trusted cyberdefenders, infrastructure providers, and select research organizations. It uses the same underlying model as Claude Fable 5 but provides lifted safeguards in approved areas for specialized high-trust use cases. The model delivers exceptional capabilities in cybersecurity, software engineering, scientific research, long-context reasoning, vision, and autonomous task execution. Anthropic initially deployed Claude Mythos 5 through Project Glasswing in collaboration with the U.S. government to help protect critical software and infrastructure. The model also shows strong potential in life sciences, including protein design, molecular biology hypothesis generation, and genomics research. Claude Mythos 5 is built for organizations that need frontier AI capabilities under controlled, trusted-access conditions.Starting Price: $10 per 1 million (input) -
11
GLM-5.3
Z.ai
GLM-5.3 is Z.ai’s frontier coding model designed for complex software engineering, long-horizon agent tasks, and advanced post-training research. The model uses the same base model as GLM-5.2, with improvements coming from scaled post-training across more environments, more diverse tasks, and larger compute investment. GLM-5.3 delivers stronger coding performance, better task ownership, improved benchmark results, and greater efficiency across realistic development workflows. It is built to handle complex coding tasks, production-style engineering work, research environments, automation tasks, and agentic workflows that require multi-step execution. The model also shows emergent cyber capabilities in vulnerability discovery and exploitation-chain reasoning, with safety evaluation and hardening planned before open-weight release.Starting Price: Free -
12
GLM-5.2
Z.ai
GLM-5.2 is an advanced AI foundation model designed to support complex reasoning, coding, and long-range agentic tasks. It helps developers, teams, and organizations build intelligent systems that can understand instructions, solve technical problems, and assist with demanding workflows. The model is especially useful for software engineering, automation, research, and productivity-focused applications. GLM-5.2 is built to handle large amounts of context, making it suitable for projects that require deeper understanding across extended conversations, documents, or codebases. Its mixture-of-experts design helps balance strong performance with more efficient model operation. GLM-5.2 gives businesses and developers a powerful AI tool for creating smarter applications, improving technical workflows, and supporting advanced digital experiences.Starting Price: Free -
13
GPT-5.6 Terra
OpenAI
GPT-5.6 Terra is a balanced model in the GPT-5.6 series designed for everyday work, coding, agentic workflows, cybersecurity support, biology analysis, and enterprise automation. It sits between GPT-5.6 Sol, the flagship model, and GPT-5.6 Luna, the faster and lower-cost option. Terra is positioned to deliver competitive performance to GPT-5.5 while being significantly cheaper to run. The model supports improved reasoning, coding, tool coordination, long-horizon workflows, and legitimate defensive security work. It is part of a model family built with layered safeguards, including trained refusals, real-time misuse classifiers, account-level review, differentiated access, monitoring, and continued red-team testing. GPT-5.6 Terra helps developers, enterprises, and technical teams access strong AI capabilities with a more practical balance of intelligence, speed, and cost.Starting Price: $2 per 1M tokens (input) -
14
Grok 4.5
SpaceXAI
Grok 4.5 is SpaceXAI’s advanced AI model built for coding, agentic tasks, engineering work, and knowledge-intensive productivity. The model is trained on coding, science, engineering, and math data, with reinforcement learning focused on multi-step software engineering and technical workflows. It is designed to handle real-world development tasks such as debugging, Rust and C/C++ work, terminal tasks, long-running agentic rollouts, and end-to-end app creation from a single prompt. Grok 4.5 is also built for fast serving, token efficiency, and lower-cost execution, with pricing based on input and output token usage. Beyond coding, the model supports business productivity tasks in Grok Build, including Excel modeling, PowerPoint diagram creation, Word writing, and research-assisted office workflows. Available through Grok Build, Cursor, and the SpaceXAI API console, Grok 4.5 gives developers and teams a high-performance model for building software, automating work, and more.Starting Price: $2 per million input tokens -
15
GPT-5.6 Luna
OpenAI
GPT-5.6 Luna is the fast and affordable model in OpenAI’s GPT-5.6 series, built to bring strong capability to users and developers who need practical intelligence with lower overhead. In the new GPT-5.6 naming system, the number identifies the model generation, while Sol, Terra, and Luna identify durable capability tiers that can advance on their own cadence, giving people and developers clearer choices across intelligence, speed, and cost. Luna sits alongside Sol, the flagship model, and Terra, the balanced model for everyday work, as part of a family designed for broader access to next-generation AI. During the limited preview, GPT-5.6 models are initially available through the API and Codex to a select group of trusted partners and organizations, with plans for broader availability in ChatGPT, Codex, and the API. OpenAI developed GPT-5.6 Sol, Terra, and Luna with its most robust safeguards to date, with configurations matched to each model’s capabilities.Starting Price: $0.20 per 1M tokens (input) -
16
Claude Fable 5
Anthropic
Claude Fable 5 is an advanced AI model from Anthropic designed to assist with software engineering, research, knowledge work, vision tasks, and complex reasoning. Built on the Mythos-class architecture, it delivers significantly improved performance across coding, analysis, and long-context workflows. The model can handle extended autonomous tasks while maintaining focus and consistency over large amounts of information. Claude Fable 5 integrates advanced reasoning, multimodal understanding, and memory capabilities to support professional and enterprise use cases. Anthropic has implemented specialized safeguards that automatically route certain high-risk cybersecurity, biology, chemistry, and model distillation requests to a different model. Claude Fable 5 helps organizations and professionals accelerate complex work while maintaining strong safety and governance controls.Starting Price: $10 per 1 million (input) -
17
Kimi K3
Moonshot AI
Kimi K3 is Moonshot AI’s most capable model, built for frontier intelligence scenarios such as software engineering, knowledge work, deep reasoning, and multimodal understanding. The model has 2.8 trillion parameters and uses Kimi Delta Attention, a hybrid linear attention mechanism, along with Attention Residuals for long-context performance. Kimi K3 supports a 1 million token context window, making it useful for analyzing large codebases, long documents, complex knowledge bases, and multi-step workflows. It includes native visual understanding for images and videos, with support for structured message formats, base64 image input, uploaded video files, and multimodal reasoning. Developers can use Kimi K3 through an OpenAI-compatible API with support for streaming, structured JSON output, partial mode, custom tools, dynamic tool loading, and automatic context caching.Starting Price: $3 per 1M tokens (input) -
18
Gemini 3.7 Flash
Google
Gemini 3.7 Flash is Google’s most intelligent workhorse model yet for coding and agents, delivering substantial improvements across software engineering, knowledge work, web development, and complex business workflows. It shows stronger performance in debugging and issue resolution, higher first-pass code accuracy, and improved generation of production-ready code. For web development, the model creates more functional layouts and feature-complete applications in fewer prompts, with strong design adherence when working from screenshots, images, or complete design systems. In knowledge-dense fields such as finance, law, and biosciences, it provides improved reasoning, accuracy, and complex-document understanding. Gemini 3.7 Flash also performs more effectively on real-world workflow automation and multimodal tasks, supporting use cases ranging from interactive web experiences and data stories to robotics and dynamically generated 3D content.Starting Price: $0.75 per 1M tokens (input) -
19
Gemini 3.6 Flash
Google
Gemini 3.6 Flash is Google’s newest Flash model built for efficient, reliable, production-scale AI agents. The model improves on Gemini 3.5 Flash with stronger coding, knowledge work, multimodal performance, computer use, and agentic workflow execution. Gemini 3.6 Flash is designed to use fewer output tokens, take fewer reasoning steps, reduce unnecessary tool calls, and lower the cost of complex AI tasks. It supports document parsing, chart analysis, data analysis, report drafting, code migrations, visual understanding, and multi-agent orchestration. The model is available through the Gemini API, Google AI Studio, Android Studio, Google Antigravity, Gemini Enterprise Agent Platform, Gemini Enterprise app, and the Gemini app. Built for developers and enterprises, Gemini 3.6 Flash helps teams build faster, lower-cost, and more capable AI agents across coding, analysis, productivity, and multimodal workloads.Starting Price: $1.50 per 1M tokens (input) -
20
Claude Sonnet 5
Anthropic
Claude Sonnet 5 is Anthropic's latest AI model, designed to deliver stronger agentic capabilities for coding, reasoning, tool use, and knowledge work while maintaining the efficiency of the Sonnet family. The model can independently plan tasks, use external tools such as browsers and terminals, and complete complex workflows that previously required larger AI models. Sonnet 5 significantly improves upon Claude Sonnet 4.6 with better reasoning, coding performance, reduced hallucinations, stronger safety behavior, and more effective autonomous task execution. It is available across Claude plans and through the Claude API with OpenAI-style developer access for application integration. Anthropic also introduced lower introductory API pricing, making Sonnet 5 a cost-effective option for developers building AI-powered products. By combining advanced agentic capabilities with improved safety and competitive pricing, Claude Sonnet 5 helps developers build more capable AI applications.Starting Price: $2 per 1M tokens (input) -
21
Qwen3.8-Max
Alibaba
Qwen3.8-Max is Qwen’s most capable model to date, built as a Max-class AI model for coding, work, research, long-horizon tasks, and multimodal agents. It scales to 2.4 trillion parameters with 95 billion active parameters and is available through QwenCloud. The model is designed to complete complex, open-ended tasks end to end with greater reliability and minimal human involvement. Qwen3.8-Max supports autonomous coding workflows, agentic development, research reproduction, visual reasoning, document understanding, video analysis, and real-world productivity tasks. It can integrate with popular agent frameworks and coding assistants, including Claude Code, Codex, Qoder CLI, Qwen Code, and OpenClaw. Built for developers, researchers, enterprises, and AI agent builders, Qwen3.8-Max helps teams automate sophisticated work across code, documents, tools, interfaces, and multimodal content.Starting Price: $2 per 1M (input) -
22
Gemini 3.5 Pro
Google
Gemini 3.5 Pro is Google’s anticipated next-generation Pro model in the Gemini 3.5 series, designed for advanced reasoning, coding, multimodal understanding, and agentic workflows. It is expected to build on Google’s Gemini 3 family with stronger performance for complex tasks that require planning, context handling, tool use, and deep problem solving. The model is aimed at users who need more power than faster Flash models for demanding development, research, automation, and enterprise AI use cases. Gemini 3.5 Pro is expected to support sophisticated workflows across text, code, files, multimodal inputs, and connected tools. Developers and organizations will likely use it through Google’s AI platforms for building assistants, agents, coding tools, analysis systems, and productivity applications. As an upcoming Pro-tier model, Gemini 3.5 Pro is positioned for high-value workloads where accuracy, reasoning quality, and advanced task execution matter more than maximum speed. -
23
MiniMax M3
MiniMax
MiniMax M3 is an open-weight multimodal AI model designed for coding, agentic workflows, long-context reasoning, and complex automation tasks. The model combines frontier-level coding performance, native multimodal understanding, and a context window of up to 1 million tokens. MiniMax M3 uses MiniMax Sparse Attention to improve long-context efficiency while reducing compute requirements for large-scale inputs. It supports text, image, and video understanding, making it useful for workflows that combine code, documents, visual references, and tool-driven tasks. The model is built for repository-scale reasoning, software engineering, autonomous task execution, tool calling, and multi-step agent workflows. MiniMax M3 helps developers, AI teams, and enterprises build capable agents that can reason across large contexts and work with multimodal information.Starting Price: $0.30 per million input tokens -
24
DeepSeek-V4-Pro
DeepSeek
DeepSeek-V4-Pro is a large-scale Mixture-of-Experts (MoE) language model designed for advanced reasoning, coding, and long-context understanding. It features 1.6 trillion total parameters with 49 billion activated parameters, enabling high performance while maintaining efficiency. The model supports an exceptionally large context window of up to one million tokens, allowing it to process extensive documents and workflows. It uses a hybrid attention architecture to optimize long-context performance and reduce computational cost. DeepSeek-V4-Pro is trained on over 32 trillion tokens, improving its knowledge and reasoning capabilities. It also includes advanced optimization techniques for stability and faster convergence during training. The model supports multiple reasoning modes, allowing users to balance speed and accuracy based on their needs. Overall, it provides a powerful open-source solution for complex AI tasks and large-scale applications.Starting Price: $0.435 per 1M tokens (input) -
25
Muse Spark 1.2
Meta
Muse Spark 1.2 is Meta’s coding-focused model update designed to power Muse Code and improve software engineering workflows. The model is built for code generation, complex debugging, codebase understanding, long-horizon development tasks, and end-to-end developer workflows. Muse Spark 1.2 was co-trained with Muse Code to improve performance inside the terminal coding agent environment. It supports planning, goal conditioning, context compaction, subagent coordination, and iterative coding workflows across large repositories. The model was trained with expanded coding compute, diverse development environments, self-improvement loops, and long-running engineering tasks. Built for AI developers and software teams, Muse Spark 1.2 helps agents plan, write, validate, debug, and optimize code with greater autonomy.Starting Price: $1.25 per 1M tokens (input) -
26
Gemini 3.5 Flash
Google
Gemini 3.5 Flash is Google’s latest frontier AI model designed to combine advanced intelligence, high-speed performance, and agentic workflow execution for developers, enterprises, and everyday users. Built as part of the Gemini 3.5 family, the model excels at coding, long-horizon reasoning, multimodal understanding, and complex multi-step automation tasks while delivering significantly faster output speeds than many competing frontier models. Gemini 3.5 Flash powers AI agents capable of planning, executing, and managing workflows such as application development, codebase maintenance, data analysis, and financial document preparation through the Antigravity harness. The model also supports rich multimodal experiences by generating interactive graphics, dynamic web interfaces, animations, and advanced visual content. Gemini 3.5 Flash is integrated across Google products including the Gemini app, Google Search AI Mode, Google Antigravity, Google AI Studio, Android Studio, and more.Starting Price: $1.50 per 1M tokens (input) -
27
GLM-5.3-Flash
Z.ai
GLM-5.3-Flash is Z.ai’s natively multimodal model in the GLM-5 series (previously previewed as Ox Alpha), designed to deliver strong coding, agentic, visual, and knowledge-work performance at relatively low inference cost. It uses 320 billion total parameters with 18 billion active parameters, along with a hybrid architecture that combines sparse and linear attention to reduce the cost of long-context processing. The model supports context lengths of up to one million tokens and was trained on a 30-trillion-token multimodal corpus. GLM-5.3-Flash can reason across text, images, documents, interfaces, dashboards, and other visual information while using that feedback to refine its own outputs. Z.ai reports substantial gains over GLM-5.2 on coding and agentic benchmarks, including DeepSWE and AutomationBench, while approaching higher-cost frontier models on several evaluations.Starting Price: $0.15 per 1M tokens (input) -
28
GPT-5.5
OpenAI
GPT-5.5 is an advanced AI model designed to handle complex, real-world tasks with greater autonomy and efficiency. It quickly understands user intent and can execute multi-step workflows such as coding, research, data analysis, and document creation with minimal guidance. Instead of requiring step-by-step instructions, GPT-5.5 plans tasks, uses tools, evaluates outputs, and continues working until completion. It excels in knowledge work, software development, and analytical problem-solving, helping users move from idea to execution faster. The model is built to operate across tools and environments, making it highly effective for modern digital workflows. With strong reasoning and persistence, GPT-5.5 enables individuals and teams to complete demanding work more efficiently and accurately.Starting Price: $5 per 1M tokens (input) -
29
Claude Opus 4.8
Anthropic
Claude Opus 4.8 is a powerful AI model from Anthropic designed to deliver stronger coding, reasoning, agentic workflows, and advanced collaboration capabilities for developers, enterprises, and AI-powered productivity tasks. The model builds on Claude Opus 4.7 with improvements across coding benchmarks, practical knowledge work, alignment, and reliability while maintaining the same pricing structure. Claude Opus 4.8 introduces enhanced honesty and reasoning behavior, making it less likely to generate unsupported claims or overlook flaws during complex tasks such as software development and agent execution. The release also includes new features such as effort control settings, fast mode for lower-cost high-speed processing, and dynamic workflows in Claude Code that allow the system to coordinate hundreds of parallel subagents for large-scale tasks.Starting Price: $5 per 1M (input) -
30
Muse Spark 1.1
Meta
Muse Spark 1.1 is a multimodal reasoning model from Meta Superintelligence Labs built for agentic tasks, coding, computer use, tool use, and multimodal understanding. The model improves on the original Muse Spark with stronger performance in planning, orchestration, long-context work, coding workflows, and external app interactions. Muse Spark 1.1 can manage a 1 million token context window, remember earlier actions, retrieve important information, compact context, and delegate tasks across parallel subagents. It is designed to operate across tools, MCP servers, custom skills, browsers, native apps, scripts, images, video, PDFs, and audio-based workflows. Developers can access Muse Spark 1.1 through the new Meta Model API public preview, while users can try it in Thinking mode in the Meta AI app and on meta.ai.Starting Price: $1.25 per 1M tokens (input)
