Lenovo’s Tianxi AI said its in-house coding agent framework, TianxiCode, working with DeepSeek-v4.1-Flash, ranked first globally on the Lite leaderboard of SWE-bench-Live with a 71% problem resolution rate. The result has passed official review and received a "Verified" label.
SWE-bench-Live evaluates a model or agent’s ability to generate code patches and fix issues based on real problems drawn from GitHub projects. The benchmark requires participants to submit full execution traces, which are then reviewed by the official team.
According to the release carried by Techub News, TianxiCode demonstrated three engineering capabilities in the evaluation: cross-file multi-hop retrieval with precise context management, self-driven planning with multi-round tool use, and closed-loop patch generation with test-driven self-repair. Lenovo said the framework will be used in its AI hardware products in the future. The item cited Quantum Bit as the source.
Lenovo’s Tianxi AI said its self-developed professional coding agent framework, TianxiCode, working with DeepSeek-v4.1-Flash, ranked first globally on the Lite leaderboard of the SWE-bench-Live software engineering benchmark with a 71% problem resolution rate. The result has also passed official review.
SWE-bench-Live review process requires full execution traces
SWE-bench-Live is built on real issues from GitHub projects and evaluates how well models and agents can generate code patches and fix problems. The evaluation process requires the submission of full-chain execution traces, followed by strict review from the official team.
TianxiCode and DeepSeek-v4.1-Flash passed that review and received a "Verified" certification.
Three engineering capabilities highlighted
Lenovo said TianxiCode showed three system engineering capabilities in the benchmark:
- Cross-file multi-hop retrieval and precise context management
- Self-driven planning and multi-round tool calling
- Closed-loop patch generation and test-driven self-healing
Lenovo said the framework will be applied to its AI hardware products in the future.
The item was attributed to Quantum Bit.
This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan. Disclaimer:
The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.
Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.
