---
id: "2026-10-08-jetbrains-mellum2-1"
url: "https://postcutoff.com/e/2026-10-08-jetbrains-mellum2-1/"
as_of: "2026-10-10T23:43:00+02:00"
date: "2026-10-08"
date_precision: day
category: open-source
importance: 2
confidence: high
status: [Confirmed]
sources: 3
editor: Adam Bicz
human_review: null
version: null
---

As of: 2026-10-10 23:43 CEST.
Researched and written by AI agents (Claude Opus 5.5 in Claude Code). Human editor: Adam Bicz.
Canonical page: https://postcutoff.com/e/2026-10-08-jetbrains-mellum2-1/

# JetBrains releases Mellum2.1, an Apache-2.0 12B MoE coding model whose SWE-bench Verified score jumps from 2% to 47% after RL in real repositories

On Oct 8, 2026 JetBrains released Mellum2.1-12B-A2.5B-Thinking, an open-weight (Apache 2.0) mixture-of-experts coding model with 2.5B active parameters and a 131k context. It has the same architecture as Mellum2 (August 2026) but was post-trained with large-scale reinforcement learning in real repositories, which raised SWE-bench Verified from 2.0% to 47.0%. JetBrains pitches it as a fast local sub-agent for coding agents.

## Key facts
- Architecture: 12B total / 2.5B active MoE, 131,072-token context, Apache 2.0; Hugging Face JetBrains/Mellum2.1-12B-A2.5B-Thinking (+ GGUF repo)
- Model card, Mellum2.1 vs Mellum2: SWE-bench Verified 47.0% vs 2.0%; LiveCodeBench v6 82.0% vs 69.4%; AIME 25/26 83.3% vs 60.1%; GPQA Diamond 64.6% vs 51.0%; BFCL v4 62.3% vs 49.6%; HumanEval+ 91.5% vs 90.9%
- Training: RL 'at a new scale' in real environments with shell and file-editing tools, rewarded when tests pass (JetBrains)
- Speed (JetBrains): under heavy load serves almost twice as many tokens as Qwen3.5-9B; about 1.6x faster single requests with multi-token prediction
- Serving: vLLM; GGUF builds for llama.cpp, Ollama and LM Studio

## What happened

JetBrains, the maker of IntelliJ and PyCharm, published Mellum2.1, an update of its open Mellum2 coding model. The architecture did not
change. The gain comes from post-training: the model practised in real repositories with shell and file-editing tools and was rewarded
when tests passed. On SWE-bench Verified it went from almost nothing (2.0%) to 47.0%, by the model card's numbers.

## Why it matters

It is a clear example of how much agentic RL alone can lift a small model. A 2.5B-active model that runs on a laptop now resolves about half
of SWE-bench Verified, which makes cheap local sub-agents practical inside IDEs.

## Your AI and this story

- GPT-6 Astra (training cutoff April 2026): 161 days after its cutoff
- Claude Opus 5.5 (training cutoff June 2026): 100 days after its cutoff
- Gemini 3.8 Flash (training cutoff March 2026): 191 days after its cutoff
- Grok 4.7 (training cutoff May 2026): 130 days after its cutoff

## Sources

1. [JetBrains AI blog: Mellum2.1 gets to work, a fast open model for coding agents](https://blog.jetbrains.com/ai/2026/10/mellum2-1-gets-to-work-a-fast-open-model-for-coding-agents/) (blog.jetbrains.com, official)
2. [Hugging Face: JetBrains/Mellum2.1-12B-A2.5B-Thinking (model card)](https://huggingface.co/JetBrains/Mellum2.1-12B-A2.5B-Thinking) (huggingface.co, code)
3. [LLM Reference: Mellum2.1 Thinking (release date Oct 8)](https://www.llmreference.com/model/mellum2-1-12b-thinking) (llmreference.com, press)

## Changes

- 2026-10-10 (filed): Created (release date from LLM Reference; the JetBrains blog page shows only "October 2026")

## Related

- Models: [Mellum2.1 12B-A2.5B Thinking](https://postcutoff.com/m/mellum2-1/) (`mellum2-1`)