# K2-Horizon-7B ![Organization](https://img.shields.io/badge/Org-IFM-555) ![Context](https://img.shields.io/badge/Context-512K%20tokens-1f6feb) ![Pricing](https://img.shields.io/badge/Price-Open%20weights-2da44e) ![License](https://img.shields.io/badge/License-Apache%202.0-d29922) ![Modality](https://img.shields.io/badge/Modality-Text-8957e5) ![Released](https://img.shields.io/badge/Released-Sep%202026-555) > **Quick answer:** K2-Horizon-7B is IFM's medium dense model — a 7B-core decoder-only model with a native 512K-token context window. It beats Gemma 4-12B, Qwen3.5-9B, and IBM Granite 4.2-8B on SWE-bench Verified (68.4%), HMMT Feb 2026 (73.3%), and Terminal-Bench 2.1 (39.1%), despite being smaller than several of its comparison models. Open weights, Apache 2.0. ## At a Glance **Where K2-Horizon-7B leads** - Strong SWE-bench Verified score (68.4%) — well ahead of larger dense peers like Gemma 4-12B (30.6%) and Qwen3.5-9B (50.8%) - Leads its comparison set on HMMT Feb 2026 competition math (73.3%) and long-context reasoning (AA-LCR, 68.0%) - 512K native context in a 7B-class model **Where it lags** - BrowseComp (59.0%) is close to but not ahead of GPT-5 (54.9%) and LongCat Flash Thinking (56.6%) — a tighter race than other benchmarks - Smaller model, so absolute scores on frontier-class benchmarks (HLE: 18.6%) remain modest in absolute terms even where relatively strong **Best for:** teams wanting strong coding/agentic capability (especially SWE-bench-style tasks) in a compact, self-hostable 7B footprint. ## What K2-Horizon-7B Is K2-Horizon-7B is the medium-sized dense model in IFM's K2-Horizon family — a 7B-core decoder-only architecture evaluated across agentic, coding, long-context, and reasoning benchmarks. It ships with intermediate checkpoints released alongside the final model, plus public training data, recipe, training code, and evaluation resources — a notably high transparency bar for a model this size. Its comparative strength on SWE-bench Verified and HMMT suggests it was tuned with real coding and competition-math workloads in mind rather than only broad instruction-following. ## Specifications | Field | Value | |---|---| | Organization | IFM | | Parameters | 7B (dense) | | Context window | 524,288 tokens (512K) | | Architecture | Dense decoder-only | | License | Apache 2.0 | | Release date | 2026-09 | | Modality | Text | ## Pricing | | Input (per 1M tokens) | Output (per 1M tokens) | |---|---|---| | Open weights | — | — | Open weights: free to download and self-host. Validated SGLang/vLLM recipe: single-GPU (TP=1). ## Context Window K2-Horizon-7B has a **524,288-token (512K) context window**, native from the midtraining stages onward — long-context reasoning (AA-LCR) scores 68.0%, ahead of same-class peers. ## Public Benchmark Scores | Benchmark | Score | Source | Date | |---|---|---|---| | [HMMT 2026](/benchmarks/hmmt/hmmt-2026) | 73.3 | [IFM model card](https://huggingface.co/IFM/K2-Horizon-7B) | 2026-09 | | [SWE-bench Verified](/benchmarks/princeton-nlp-openai/swe-bench-verified) | 68.4 | [IFM model card](https://huggingface.co/IFM/K2-Horizon-7B) | 2026-09 | | [Humanity's Last Exam](/benchmarks/scale-ai/humanitys-last-exam) | 18.6 | [IFM model card](https://huggingface.co/IFM/K2-Horizon-7B) | 2026-09 | | [SciCode](/benchmarks/scicode-team/scicode) | 31.6 | [IFM model card](https://huggingface.co/IFM/K2-Horizon-7B) | 2026-09 | | [AA-LCR](/benchmarks/artificial-analysis/aa-lcr) | 68.0 | [IFM model card](https://huggingface.co/IFM/K2-Horizon-7B) | 2026-09 | | [TerminalBench 2.1](/benchmarks/tbench/terminalbench-21) | 39.1 | [IFM model card](https://huggingface.co/IFM/K2-Horizon-7B) | 2026-09 | | [τ³ Banking](/benchmarks/tau-bench/tau3-banking) | 25.8 | [IFM model card](https://huggingface.co/IFM/K2-Horizon-7B) | 2026-09 | | [BrowseComp](/benchmarks/openai/browsecomp) | 59.0 | [IFM model card](https://huggingface.co/IFM/K2-Horizon-7B) | 2026-09 | *Scores above are reported by IFM and shown for context; not Benchgen measurements. BrowseComp: IFM used the Discard-all@95k context-length protocol from the DeepSeek-V3.2 technical report; comparison models may use different harnesses.* ## Use K2-Horizon-7B via API ```python from openai import OpenAI client = OpenAI(base_url="http://localhost:30000/v1", api_key="EMPTY") response = client.chat.completions.create( model="IFM/K2-Horizon-7B", messages=[{"role": "user", "content": "Explain the result step by step."}], temperature=1.0, top_p=0.95, max_tokens=32768, extra_body={"chat_template_kwargs": {"reasoning_effort": "high"}}, ) print(response.choices[0].message.content) ``` ## Frequently Asked Questions

What is K2-Horizon-7B?

It's IFM's medium dense model in the K2-Horizon family: 7B parameters, 512K-token context, with strong SWE-bench Verified and HMMT scores relative to its size.

What is K2-Horizon-7B's context window?

524,288 tokens (512K).

How much does K2-Horizon-7B cost?

Open weights — free to download; cost is hosting/inference only.

Is K2-Horizon-7B open source?

Yes, released under the Apache 2.0 license.

--- *Specs and scores sourced from IFM's official Hugging Face model card; third-party benchmark scores attributed inline. Last updated 2026-09-03.*

K2 Horizon 7B — Evaluations

Evaluations