Crata/GRM-3.2-Sky-FP8

• 截至2026年9月1日，本模型的fp8量化仍然在400B以下的开源模型中显得相当出色，值得替代qwen3.8 27b或deepseek v4flash本地部署。

• 配置文件支持yarn开启4倍上下文，支持到1m

• 参考的质量测试：https://wangxian001.github.io/SQL_LLM_benchmark/

  • 在上述多轮工具调用SQL测试中，本模型fp8权重、fp8kv、yarn情况下的评分甚至略高于qwen3.8 27b官方fp8权重、fp8kv、不开启yarn

• As of September 1, 2026, the FP8 quantization of this model still stands out among open-source models with less than 400B parameters, making it a viable alternative to Qwen3.8 27B or DeepSeek V4Flash for local deployment.

• The model supports Yarn to extend the context length up to 4 times, with a maximum context length of 1 million tokens.

• For reference, quality benchmark can be found at: https://wangxian001.github.io/SQL_LLM_benchmark/

  • In the multi-turn SQL benchmark test mentioned above, the model's performance with FP8 weights, FP8 KV cache, and Yarn support even slightly outperforms the official FP8 weight, FP8 KV cache, and no Yarn version of Qwen3.8 27B.

启动指令 RUN（vllm）

VLLM_ALLOW_LONG_MAX_MODEL_LEN=1 VLLM_USE_FASTOKENS=1 CUDA_VISIBLE_DEVICES=0 CUDA_DEVICE_ORDER=PCI_BUS_ID vllm serve /mnt/xg8/bigmodel/runllm/all35B/GRM-3.2-Sky-FP8 \
  --gpu-memory-utilization 0.9 \
  --kv-cache-dtype fp8 \
  --max-model-len 1010000 \
  --max-num-seqs 8 \
  --max-num-batched-tokens 16384 \
  --enable-chunked-prefill \
  --enable-prefix-caching \
  --async-scheduling \
  --enable-auto-tool-choice --tool-call-parser qwen3_coder \
  --reasoning-parser qwen3 \
  --served-model-name GRM-3.2-Sky-FP8 \
  --host 0.0.0.0 --port 8000

1. Introduction

We're introducing GRM-3.2-Sky, our latest flagship model built for long-horizon agentic tasks and extremely difficult reasoning problems. GRM-3.2-Sky marks a substantial leap in long-horizon task capability over its predecessor, GRM-2.6-Plus, and is designed to serve as a dependable engine for complex, multi-step workflows.

The model is purpose-built for long-horizon agentic tasks and problems that are simply hard — difficult coding challenges, advanced mathematics, and rigorous logical reasoning. GRM-3.2-Sky aims to sustain coherent, goal-directed behavior over extended interactions, making it well suited for users who need a model that doesn't lose the thread across many steps of tool use, planning, and self-correction.

2. Key Capabilities

• Long-Horizon Agentic Mastery: GRM-3.2-Sky is specifically optimized to maintain coherence, planning quality, and task fidelity across long, multi-step agentic workflows, a significant step up from GRM-2.6-Plus.
• Elite Reasoning on Hard Problems: Strong performance on difficult coding, advanced mathematics, and logical reasoning tasks, with careful, structured step-by-step problem-solving.
• Robust Coding Ability: Handles complex, difficult codebases and multi-file coding tasks, including debugging, refactoring, and long-running terminal/agentic coding sessions.
• Consistent Logical Reasoning: Built to reason carefully through multi-constraint logic problems without losing track of intermediate steps.
• Flagship-Class Performance: Positioned as the top of the GRM lineup, intended to compete head-to-head with frontier-scale models on the hardest tasks.

3. Performance

GRM-3.2-Sky is designed as our most capable model to date for long-horizon agentic work and difficult reasoning. It builds directly on the strengths of GRM-2.6-Plus while specifically targeting the failure modes that emerge over long task horizons — drift, inconsistency, and loss of goal state — resulting in meaningfully improved reliability across extended sessions.

Its core strength is sustained intelligence over time: elite-level reasoning, resilient long-horizon planning, and the ability to stay on task through difficult, multi-step coding, math, and logic problems.

Detailed Benchmarks

	GRM-3.2-Sky	GPT-5.6-Luna	Sonnet 5	Qwen3.6-27B	Muse-Glimmer-30B	DeepSeek-V4-Flash
Knowledge & STEM
Multidisciplinary knowledge MMLU-Pro	89.5	—	—	86.2	—	86.2
General knowledge MMLU-Redux	96.9	—	—	93.5	—	—
Scientific reasoning GPQA Diamond	90.6	92.3	—	87.8	83.5	88.1
Reasoning & Coding
Competitive coding LiveCodeBench v6	87.7	—	—	83.9	—	91.6
Competition mathematics HMMT Feb 26	86.4	—	—	84.3	—	94.8
Mathematical reasoning AIME26	96.3	—	—	94.1	94.7	95.8
General Agent
Agentic coding SWE-bench Verified	81.4	—	85.2	77.2	76.0	79.0
Real-world software engineering SWE-bench Pro	58.3	62.7	63.2	53.5	51.2	52.6
Agentic terminal coding Terminal-Bench 2.1 (Terminus-2)	66.3	84.7	80.4	59.3	51.7	56.9
Repo-level code generation NL2Repo	35.6	—	—	36.2	—	—
Long-horizon agentic coding DeepSWE	19.5	67.2	53.8	—	—	53.0

Scores are taken from each provider's own published model card, blog post, or system card where available; "—" indicates a score was not publicly reported by that provider at the time of writing. Different labs may use different agent scaffolds when reporting SWE-bench and Terminal-Bench results, so cross-provider comparisons should be read with that caveat.

4. Family

The GRM-3.2 family is available in various sizes to suit every use case.

Model	Size	Domain
GRM-3.2-Sky	35B-A3B	Flagship model for long-horizon tasks
GRM-3.2-Cliff	9B	Capable model for low GPU environments
GRM-3.2-Turf	1.2B	Lightweight model for practical reasoning

5. Architecture

GRM-3.2-Sky is built on the Ornith-1.0-35B architecture, a 35B-parameter Mixture-of-Experts model with ~3B active parameters (35B-A3B), optimized for long-horizon agentic workflows, difficult coding, advanced mathematics, and rigorous logical reasoning, while remaining efficient to deploy thanks to its sparse activation.

---

GRM-3.2-Sky is developed by OrionLLM and released under the Apache 2.0 License.