JetBrains has released Mellum2.1, an open-source coding model positioned as an upgrade to Mellum2, which the company introduced in June this year. The new version keeps the same 12 billion total parameters and activates 2.5 billion parameters per generation, but JetBrains said it performs better at identifying code issues, editing files, and checking results. The company said it did not change the model architecture and instead improved performance mainly through reinforcement learning. According to JetBrains, the team ran millions of sandbox tasks across thousands of environments during training, with the model using terminal and file-editing tools inside real code repositories and receiving rewards when tests passed. In JetBrains’ own benchmarks, Mellum2.1 raised its success rate on SWE-bench Verified from 2% to 47%, slightly below Qwen3.5-9B’s 50%. On LiveCodeBench v6, however, Mellum2.1 scored 82%, ahead of Qwen3.5-9B’s 75.4%. JetBrains also said that under heavy-load inference on a single H200, the new model’s output throughput was close to twice that of Qwen3.5-9B. Model weights and a GGUF quantized version are now available on Hugging Face under an Apache 2.0 license for local deployment.
JetBrains has released Mellum2.1 as an open-source coding model. Compared with Mellum2, which the company introduced in June this year, the new version keeps the same 12 billion total parameters and activates 2.5 billion parameters per generation, while improving its ability to spot code issues, modify files, and verify results.
JetBrains said it did not alter the model architecture. The gains came mainly from reinforcement learning. During training, the team ran millions of sandbox tasks across thousands of environments. The model used terminal and file-editing tools inside real code repositories and received rewards when it successfully passed tests.
In JetBrains’ internal testing, Mellum2.1 lifted its success rate on SWE-bench Verified, a benchmark for fixing real software issues, from 2% in the previous version to 47%. That result was slightly below Qwen3.5-9B at 50%.
On LiveCodeBench v6, Mellum2.1 posted a score of 82%, ahead of Qwen3.5-9B’s 75.4%.
JetBrains also said the new model delivered output throughput close to twice that of Qwen3.5-9B when running under heavy load on a single H200.
The model weights and a GGUF quantized version are now available on Hugging Face under an Apache 2.0 license, allowing developers to deploy the model locally.
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.
