cafe3310 commited on
Commit
95feb31
·
verified ·
1 Parent(s): a4ccd6b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -3
README.md CHANGED
@@ -12,7 +12,7 @@ We're introducing **Ling-3.1-flash**, an upgraded native hybrid reasoning model
12
 
13
  With stronger real-world capabilities, Ling-3.1-flash shows improvements across a range of real-world tasks, including general-purpose agents, search, office productivity, and software engineering. It demonstrates strong performance on benchmarks such as **GDPVal-AA V2.1, HealthBench Professional, FrontierSWE, and Terminal-Bench 4.0**, covering office tasks, healthcare, software engineering, and terminal-based tool use. The model also shows improvements in specialized domains such as finance and materials science.
14
 
15
- <img src="https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1791987400376-c61f96e8-41a7-4377-80c2-173a01132dc5.png" alt="Ling-3.1-flash performance overview">
16
 
17
  ## Model Overview
18
  The table and diagram below summarize the model’s specifications and architecture.
@@ -31,14 +31,14 @@ The table and diagram below summarize the model’s specifications and architect
31
  | Dense Intermediate Size | 10240 |
32
  | Vocabulary Size | 157184 |
33
 
34
- <img src="https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1791966963600-8408d222-3dc1-43c6-87af-f4f905a681db.png" width="600" alt="Ling-3.1-flash model architecture">
35
 
36
  ## Evaluation
37
  We have conducted a comprehensive evaluation of Ling-3.1-flash across multiple authoritative benchmarks. **Ling-3.1-flash** performs strongly on representative coding/agentic benchmarks including **Terminal-Bench 4.0**, **SWE-bench Pro**, **FrontierSWE**, **GDPVal-AA V2.1**, **Automationbench**, **MCP-Atlas**, and **SkillsBench**, among others. In practice, Ling-3.1-flash delivers a strong user experience across **Claude Code**, **Kilo Code**, **Qwen Code**, **Hermes Agent**, **OpenClaw**, and other frameworks. Beyond agentic tasks, Ling-3.1-flash performs strongly on real-world tasks across **STEM**, **healthcare**, and **finance**.
38
 
39
  We also introduce **LingCode**, a harness built on the open-source coding agent Pi and enhanced to support complex, long-horizon coding tasks that run for hours, involve hundreds of rounds of tool calls, and require repeated context compaction. LingCode delivers significant performance gains on some coding agent tasks, and we are considering open-sourcing it in the future.
40
 
41
- <img src="https://intranetproxy.alipay.com/skylark/lark/0/2026/png/202156655/1791988674787-5f49370c-0853-4940-96ea-9f1e4ff20d57.png" alt="Ling-3.1-flash evaluation results">
42
 
43
  ## Quickstart
44
  ### SGLang (Recommended)
 
12
 
13
  With stronger real-world capabilities, Ling-3.1-flash shows improvements across a range of real-world tasks, including general-purpose agents, search, office productivity, and software engineering. It demonstrates strong performance on benchmarks such as **GDPVal-AA V2.1, HealthBench Professional, FrontierSWE, and Terminal-Bench 4.0**, covering office tasks, healthcare, software engineering, and terminal-based tool use. The model also shows improvements in specialized domains such as finance and materials science.
14
 
15
+ ![image](https://cdn-uploads.huggingface.co/production/uploads/678143d4ddb1f1ce16daca8e/D6Ns_px2xy0ns6l4kXWVF.png)
16
 
17
  ## Model Overview
18
  The table and diagram below summarize the model’s specifications and architecture.
 
31
  | Dense Intermediate Size | 10240 |
32
  | Vocabulary Size | 157184 |
33
 
34
+ ![image](https://cdn-uploads.huggingface.co/production/uploads/678143d4ddb1f1ce16daca8e/qwmfVu_K08auI7R_ijbIX.png)
35
 
36
  ## Evaluation
37
  We have conducted a comprehensive evaluation of Ling-3.1-flash across multiple authoritative benchmarks. **Ling-3.1-flash** performs strongly on representative coding/agentic benchmarks including **Terminal-Bench 4.0**, **SWE-bench Pro**, **FrontierSWE**, **GDPVal-AA V2.1**, **Automationbench**, **MCP-Atlas**, and **SkillsBench**, among others. In practice, Ling-3.1-flash delivers a strong user experience across **Claude Code**, **Kilo Code**, **Qwen Code**, **Hermes Agent**, **OpenClaw**, and other frameworks. Beyond agentic tasks, Ling-3.1-flash performs strongly on real-world tasks across **STEM**, **healthcare**, and **finance**.
38
 
39
  We also introduce **LingCode**, a harness built on the open-source coding agent Pi and enhanced to support complex, long-horizon coding tasks that run for hours, involve hundreds of rounds of tool calls, and require repeated context compaction. LingCode delivers significant performance gains on some coding agent tasks, and we are considering open-sourcing it in the future.
40
 
41
+ ![image](https://cdn-uploads.huggingface.co/production/uploads/678143d4ddb1f1ce16daca8e/VNCYlkUqCCY-QN-_dAmy0.png)
42
 
43
  ## Quickstart
44
  ### SGLang (Recommended)