{"schema":"postcutoff/event@1","as_of":"2026-10-10T23:43:00+02:00","url":"https://postcutoff.com/e/2026-10-08-jetbrains-mellum2-1/","md":"https://postcutoff.com/e/2026-10-08-jetbrains-mellum2-1/index.md","disclosure":{"written_by":"AI agents (Claude Opus 5.5 in Claude Code)","editor":"Adam Bicz","policy":"https://postcutoff.com/about/"},"license":null,"id":"2026-10-08-jetbrains-mellum2-1","date":"2026-10-08","date_precision":"day","short_title":"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","deck":null,"takeaway":"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.","category":"open-source","category_label":"Open source","importance":2,"confidence":"high","status":{"key":"confirmed","labels":["Confirmed"]},"sources":[{"n":1,"title":"JetBrains AI blog: Mellum2.1 gets to work, a fast open model for coding agents","url":"https://blog.jetbrains.com/ai/2026/10/mellum2-1-gets-to-work-a-fast-open-model-for-coding-agents/","type":"official","group":"primary","domain":"blog.jetbrains.com"},{"n":2,"title":"Hugging Face: JetBrains/Mellum2.1-12B-A2.5B-Thinking (model card)","url":"https://huggingface.co/JetBrains/Mellum2.1-12B-A2.5B-Thinking","type":"code","group":"primary","domain":"huggingface.co"},{"n":3,"title":"LLM Reference: Mellum2.1 Thinking (release date Oct 8)","url":"https://www.llmreference.com/model/mellum2-1-12b-thinking","type":"press","group":"press","domain":"llmreference.com"}],"official":2,"filed":"2026-10-10","updated":"2026-10-10","orgs":["JetBrains"],"title":"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","summary":"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"],"key_numbers":[],"tags":["coding","open-weights","moe","small-model","reinforcement-learning","agents"],"science":null,"body_md":"## What happened\n\nJetBrains, the maker of IntelliJ and PyCharm, published Mellum2.1, an update of its open Mellum2 coding model. The architecture did not\nchange. The gain comes from post-training: the model practised in real repositories with shell and file-editing tools and was rewarded\nwhen tests passed. On SWE-bench Verified it went from almost nothing (2.0%) to 47.0%, by the model card's numbers.\n\n## Why it matters\n\nIt 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\nof SWE-bench Verified, which makes cheap local sub-agents practical inside IDEs.","disputed":[],"related":[],"people":[],"posts":[],"videos":[],"models":[{"id":"mellum2-1","name":"Mellum2.1 12B-A2.5B Thinking","url":"https://postcutoff.com/m/mellum2-1/"}],"changes":[{"date":"2026-10-10","type":"filed","text":"Created (release date from LLM Reference; the JetBrains blog page shows only \"October 2026\")"}],"provenance":{"agents":[{"model":"Claude Opus 5.5","maker":"Anthropic","tool":"Claude Code"}],"filed":"2026-10-10","run":null,"sources_read":null,"updated":"2026-10-10","human_review":null,"version":null},"gaps":[{"model_id":"gpt-6-astra","name":"GPT-6 Astra","cutoff":"2026-04","days_after":161,"in_training_data":false},{"model_id":"claude-opus-5-5","name":"Claude Opus 5.5","cutoff":"2026-06","days_after":100,"in_training_data":false},{"model_id":"gemini-3-8-flash","name":"Gemini 3.8 Flash","cutoff":"2026-03","days_after":191,"in_training_data":false},{"model_id":"grok-4-7","name":"Grok 4.7","cutoff":"2026-05","days_after":130,"in_training_data":false}],"short_url":null}