--- title: TeleChat2-3B canonical_url: "https://www.modelscope.cn/models/Qualcomm-AI-Hub/TeleChat2-3B" md_url: "https://www.modelscope.cn/models/Qualcomm-AI-Hub/TeleChat2-3B.md" repository: Qualcomm-AI-Hub/TeleChat2-3B last_updated: 2026-10-16 license: apache-2.0 pipeline_tag: text-generation tasks: - text-generation downloads: 0 stars: 0 tags: - llm - generative_ai - android --- # TeleChat2-3B > TeleChat2-3B - Qualcomm-AI-Hub 在 ModelScope 开源的模型。TeleChat2-3B: Optimized for Qualcomm Devices Qualcomm-AI-Hub/TeleChat2-3B 是 ModelScope 魔搭社区上的text-generation模型，采用 apache-2.0 许可。 - **Repository**: Qualcomm-AI-Hub/TeleChat2-3B - **License**: apache-2.0 - **Tasks**: text-generation - **Tags**: llm, generative_ai, android - **Downloads**: 0 - **Stars**: 0 - **Last updated**: 2026-10-16 Source: https://www.modelscope.cn/models/Qualcomm-AI-Hub/TeleChat2-3B --- ![](https://china-qaihub-public-assets.s3.cn-northwest-1.amazonaws.com.cn/qai-hub-models/models/telechat2_3b/web-assets/model_demo.png) # TeleChat2-3B: Optimized for Qualcomm Devices For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to https://github.com/Tele-AI/TeleChat2 This is based on the implementation of TeleChat2-3B found [here](https://www.modelscope.cn/models/TeleAI/TeleChat2-3B). This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models-cn/blob/main/src/qai_hub_models/models/telechat2_3b) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary). Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.cn) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) to run these models on a hosted Qualcomm® device. ## Deploying TeleChat2-3B on-device Please follow the [LLM on-device deployment](https://github.com/qualcomm/ai-hub-apps/tree/main/tutorials/llm_on_genie) tutorial. ## Getting Started Below are pre-exported model assets ready for deployment. | Runtime | Precision | Chipset | SDK Versions | Download | |---|---|---|---|---| | GENIE | w8 | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.45 | [Download](https://china-qaihub-public-assets.s3.cn-northwest-1.amazonaws.com.cn/qai-hub-models/models/telechat2_3b/releases/v0.62.1/telechat2_3b-genie-w8-qualcomm_snapdragon_8_elite_gen5.zip) | GENIE | w8 | Snapdragon® 8 Elite Mobile | QAIRT 2.45 | [Download](https://china-qaihub-public-assets.s3.cn-northwest-1.amazonaws.com.cn/qai-hub-models/models/telechat2_3b/releases/v0.62.1/telechat2_3b-genie-w8-qualcomm_snapdragon_8_elite.zip) | GENIE | w8 | Snapdragon® X Elite | QAIRT 2.45 | [Download](https://china-qaihub-public-assets.s3.cn-northwest-1.amazonaws.com.cn/qai-hub-models/models/telechat2_3b/releases/v0.62.1/telechat2_3b-genie-w8-qualcomm_snapdragon_x_elite.zip) | GENIE | w8 | Qualcomm® Dragonwing™ IQ-9075 | QAIRT 2.45 | [Download](https://china-qaihub-public-assets.s3.cn-northwest-1.amazonaws.com.cn/qai-hub-models/models/telechat2_3b/releases/v0.62.1/telechat2_3b-genie-w8-qualcomm_qcs9075.zip) | GENIE | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.39 | [Download](https://china-qaihub-public-assets.s3.cn-northwest-1.amazonaws.com.cn/qai-hub-models/models/telechat2_3b/releases/v0.62.1/telechat2_3b-genie-w8a16-qualcomm_snapdragon_8_elite_gen5.zip) | GENIE | w8a16 | Snapdragon® 8 Elite Mobile | QAIRT 2.39 | [Download](https://china-qaihub-public-assets.s3.cn-northwest-1.amazonaws.com.cn/qai-hub-models/models/telechat2_3b/releases/v0.62.1/telechat2_3b-genie-w8a16-qualcomm_snapdragon_8_elite.zip) For more device-specific assets and performance metrics, visit **[TeleChat2-3B on Qualcomm® AI Hub](https://aihub.qualcomm.cn/models/telechat2_3b)**. ## Model Details **Model Type:** Model_use_case.text_generation **Model Stats:** - Information about the model parts: Prompt Processor and Token Generator are split into 7 parts each. Each corresponding Prompt Processor and Token Generator part share weights. - Input sequence length for Prompt Processor: 128 - Maximum context length: 4096 - Minimum QNN SDK version required: 2.45 - Num of key-value heads: 8 for KV, 16 for Q - Precision: w8, w8a16 is supported - Prompt processor input (other parts): 128 tokens + KVCache initialized with pad token - Prompt processor input (part1): 128 tokens - Prompt processor output (other parts): 128 output tokens + KVCache for token generator - Prompt processor output (part1): Embeddings output - Response Rate: Rate of response generation after the first response token. - TTFT: Time To First Token is the time it takes to generate the first response token. This is expressed as a range because it varies based on the length of the prompt. The lower bound is for a short prompt (up to 128 tokens, i.e., one iteration of the prompt processor) and the upper bound is for a prompt using the full context length (4096 tokens). - Token generator input (other parts): 1 input token + past KVCache - Token generator input (part1): 128 tokens - Token generator output (other parts): 1 output token + KVCache for next iteration - Token generator output (part1): Embeddings output - Use: Initiate conversation with prompt-processor and then token generator for subsequent iterations. ## Performance Summary | Model | Runtime | Precision | Chipset | Context Length | Response Rate (tokens per second) | Time To First Token (range, seconds) |---|---|---|---|---|---|--- | TeleChat2-3B | GENIE | w8 | Snapdragon® 8 Elite Mobile | 4096 | 8.704 | 1.4257053333333334 - 45.62257066666667 | TeleChat2-3B | GENIE | w8 | Qualcomm® Dragonwing™ IQ-9075 | 4096 | 5.690698146820068 | 0.8503005 - 27.209616 | TeleChat2-3B | GENIE | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 4096 | 21.33622 | 0.07334433333333333 - 2.3470186666666666 | TeleChat2-3B | GENIE | w8a16 | Snapdragon® 8 Elite Mobile | 4096 | 19.64214 | 0.07196733333333333 - 2.3029546666666665 ## License * The license for the original implementation of TeleChat2-3B can be found [here](https://www.modelscope.cn/models/TeleAI/TeleChat2-3B). ## References * [Technical Report of TeleChat2, TeleChat2.5 and T1](https://arxiv.org/abs/2507.18013) * [Source Model Implementation](https://www.modelscope.cn/models/TeleAI/TeleChat2-3B) ## Community * For questions or feedback, chat with our engineers in [Qualcomm® AI Hub Workbench](https://workbench.aihub.qualcomm.cn). ## Usage and Limitations This model may not be used for or in connection with any of the following applications: - Accessing essential private and public services and benefits; - Administration of justice and democratic processes; - Assessing or recognizing the emotional state of a person; - Biometric and biometrics-based systems, including categorization of persons based on sensitive characteristics; - Education and vocational training; - Employment and workers management; - Exploitation of the vulnerabilities of persons resulting in harmful behavior; - General purpose social scoring; - Law enforcement; - Management and operation of critical infrastructure; - Migration, asylum and border control management; - Predictive policing; - Real-time remote biometric identification in public spaces; - Recommender systems of social media platforms; - Scraping of facial images (from the internet or otherwise); and/or - Subliminal manipulation