Tencent Unveils Its Most Powerful AI Model at Extremely Low Cost

Tencent has launched Hy4 Preview, a low-cost, open-weight AI model designed to compete with leading systems from OpenAI, Anthropic, DeepSeek, Zhipu AI, and Moonshot AI. Unlike proprietary models such as GPT-5.6 and Claude Opus 5, Hy4 Preview offers publicly available model weights that developers can download and deploy independently. It also comes with relatively affordable API pricing, starting at $0.042 per million cached input tokens, while regular input costs $0.834 and output costs $2.501 per million tokens. The model has 770 billion total parameters, activates 49 billion parameters per token, and supports a one-million-token context window. Tencent says it delivers significant improvements in software engineering, office productivity, game development, and scientific research. According to Tencent's published benchmarks, Hy4 Preview outperforms several competing frontier models in selected coding and productivity tests. However, Claude Opus 5, GPT-5.6 Sol, and other rivals remain ahead in some areas. Released on August 28, 2026, Hy4 Preview is available through Tencent Cloud, OpenRouter, and Tencent's AI applications. Developers can also download its model weights through supported open-source repositories. Hy4 Preview Specifications Hy4 Preview uses a Mixture-of-Experts (MoE) architecture, which activates only part of the model's total parameters for each token. The model contains 78 layers, including 77 MoE layers with 256 routed experts and one shared expert per layer. It also uses Gated DeepSeek Sparse Attention and identity Hyper-Connections to improve information processing. Tencent says it increased model size, context length, and training data while expanding both pre-training and post-training. The company describes Hy4 Preview as its largest generation-over-generation improvement so far. Hy4 Preview vs Major Models Tencent evaluated Hy4 Preview against several leading AI models using software engineering, agent-based, and reasoning benchmarks. The comparisons include Claude Opus 5, GPT-5.6 Sol, DeepSeek V4 Pro, GLM 5.3, and Kimi K3. Higher scores indicate better performance on each benchmark. Some rivals have two reported scores under different evaluation configurations. These are retained as published. The comparisons are Tencent-reported results, not an independent ranking. Hy4 achieved 82.9 on SWE-bench Multilingual, exceeding the listed scores for DeepSeek V4 Pro, GLM 5.3, Kimi K3, and GPT-5.6 Sol. However, Claude Opus 5 performed better on that test. Hy4 also achieved a score of 64.0 on SWE Atlas Codebase Q&A, outperforming all five competitors shown in the table. On Terminal-Bench 2.1, it scored 85.4, although several competitors reported higher results. These figures indicate that Hy4 is competitive across demanding programming tasks, rather than the strongest model in every category. Tencent Hy4 vs Hy3: Improvements Tencent also compared its new model with the previous Hy3 generation. The results show improvements in coding, tool use, research, and long-context processing. One of the largest improvements appeared on DeepSWE, where the score increased from 28.0 to 64.3. The model also improved from 30.8 to 64.0 on SWE Atlas Codebase Q&A and from 3.0 to 17.5 on ProgramBench. Tencent says these gains resulted from larger-scale training and greater use of real-world engineering tasks. Hy4 Performance The model also performed competitively on benchmarks designed to measure its ability to search information, work with documents, and use external tools. The results show that Hy4 improved significantly over Hy3 in document-heavy tasks and multi-step workflows. On JobBench, its score rose from 34.6 to 61.7, while its OneMillionBench result improved from 51.5 to 65.4. However, GLM 5.3 and Kimi K3 remained ahead in some agent and workspace evaluations. Tests Against GLM and Kimi Tencent conducted a separate blind evaluation involving 163 internal experts and 203 software engineering tasks. Participants assessed outputs produced by Hy4 Preview, GLM 5.3, and Kimi K3. Hy4 received an average score of 2.99 out of 4, compared with 2.92 for GLM 5.3 and 2.94 for Kimi K3. Hy4 performed slightly better in both comparisons, although the differences in average scores were small. The tests were conducted internally by Tencent, so they should not be treated as independent confirmation of overall model superiority. Hy4 Is Designed for Coding, Office Work, Games and Research Tencent developed Hy4 Preview with input from software engineers, game developers, financial analysts, and security specialists. The company used their work to create training data focused on practical tasks. In software engineering, Tencent says Hy4 can handle longer development tasks while improving front-end design and interaction quality. The company's 3D viewer demonstration followed specific requirements, including a three-column layout, paper-style background, serif typography, and interactive model controls. For office work, Tencent demonstrated a workflow involving 24 reimbursement claims. The model was required to examine several policy versions, staff records, budgets, invoices, and emails before preparing decisions and explanations. In game development, Hy4 created a cartoon adventure game featuring a penguin exploring a snowfield, collecting berries, opening chests, and fighting monsters. The prototype also included menus, animations, inventory management, quest information, music, and sound effects. Hy4 AI Tools for Scientific Research Tencent also reported improvements in scientific research, particularly in AI research, molecular dynamics, condensed matter physics, and pure mathematics. In one experiment, Hy4 managed several Codex sessions simultaneously while working on a small-model post-training task. The model evaluated results, adjusted research directions, and coordinated experiments targeting different performance measures. Tencent says Hy4 outperformed Codex working independently across all eight benchmarks included in that experiment. This suggests progress in AI-assisted research management, although the demonstration does not establish that Hy4 can independently develop new frontier AI systems. Tencent Hy4 Preview API Pricing Tencent has introduced usage-based API pricing for Hy4 Preview. Cached input is significantly cheaper than standard input because previously processed information can be reused. The company says its pricing is intended to make the model accessible for developers and businesses handling large workloads. Where to Access Hy4 Preview Tencent has made Hy4 available through cloud platforms, consumer applications, and open-source repositories. Tencent provides both the standard Hy4 Preview model and an FP8 version. Developers can deploy the model using supported tools such as vLLM and SGLang. The model is released under the Apache 2.0 license. Tencent also offered two weeks of free Hy4 Preview access through CodeBuddy and WorkBuddy following its launch. The company separately extended free access to Hy3 through September 30. Acknowledged Limitations in Hy4 Preview Despite the benchmark improvements, Tencent says the model still has problems that require further development. The company identified two main limitations: spending too much time reasoning through complicated tasks and repeatedly checking its own work. These behaviors can increase response times and affect efficiency. Tencent says it released the preview version early to collect feedback and identify issues. The company plans to continue improving the model through additional training and collaboration with internal teams and Tencent products. Hy4 Preview represents a substantial improvement over Hy3 and performs competitively against several frontier AI systems on selected benchmarks. However, its results do not establish that it outperforms every rival, and independent testing remains important for evaluating its performance across real-world tasks. Stay Connected with ProPakistani Get the latest tech news, telecom insights, and product launches wherever you prefer. See more ProPakistani stories in Google Search and Top Stories.