Bad Theory Labs has released BTL-3, a 27-billion parameter open-weight model designed for agentic coding and structured tool use. This model is a post-trained version of Qwen3.6-27B, offering strong performance on coding benchmarks like HumanEval with a 95.1% pass rate. BTL-3 also features a large context window of 262,144 tokens and is available under the Apache-2.0 license, with a compact edition optimized for local inference. AI
IMPACT This release provides a new open-weight model optimized for coding and tool use, potentially enhancing agentic capabilities and local inference.
RANK_REASON Release of an open-weight model based on a prior model, with benchmark results. [lever_c_demoted from research: ic=1 ai=1.0]
Read on Hugging Face Trending Models →
- Apache-2.0
- Bad Theory Labs
- badtheorylabs/BTL-3
- BFCL v4 AST
- HumanEval
- LiveCodeBench v6
- Qwen/Qwen3.6-27B
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
