I
inclusionAI: Ling-2.6-1T
● Excellent8.2
/ 10
Ling-2.6-1T is an instant (instruct) model from inclusionAI and the company’s trillion-parameter flagship, designed for real-world agents that require fast execution and high efficiency at scale. It uses a “fast...
Specifications
| Attribute | Value |
|---|---|
| Lab | Inclusionai |
| Tags | Fast Agentic Coding |
| Overall Score | 8.2/10 |
| Release Date | 2026-04 |
| Context Window | 262,144 tokens |
| Input Price / 1M | $0.07 |
| Output Price / 1M | $0.62 |
| Input Modalities | Text |
| Output Modalities | Text |
Strengths
- 1T-parameter instant model with fast execution
- 63B active parameters for efficiency
- Optimized for production agent deployments
- Strong real-time performance
Weaknesses
- Below Ring-2.6-1T on deep reasoning
- Newer lab with less ecosystem maturity
- Limited benchmark data
Best For
- Production agent systems
- Real-time coding assistance
- High-throughput reasoning
Sources & Further Reading
Related Models
Scores are aggregated from public benchmarks (MMLU, HumanEval, MATH, GSM8K, LMSYS) and normalized to a 1–10 scale. Methodology →