Ex0bit

MiniMax-SLURPY

README

License: modified-mit

What SLURPY inherits

SLURPY's weights are a forensically-driven interpolation of two complementary parents. The merge schedule is derived from a full-model scan of all 96,103 tensor pairs, targeting each tensor's interpolation ratio to the empirically measured delta between the parents.

From M2.5 — the architect

M2.5 is the foundation-builder: strong on greenfield engineering, deep reasoning, and research-grade benchmarks.

Table with columns: Benchmark, M2.5 Published
Benchmark	M2.5 Published
SWE-Bench Verified	80.2%
BrowseComp (with context mgmt)	76.3%
Multi-SWE-Bench	51.3%
AIME 2025	86.3
GPQA Diamond	85.2
SciCode	44.4
IFBench	70.0
HLE (w/o tools)	19.4
GDPval-MM (office work)	59.0% avg win rate

From M2.7 — the operator

M2.7 is the execution specialist: RL-tuned for multi-step tool use, terminal ops, agentic scaffolding, and production-grade software engineering.

Table with columns: Benchmark, M2.7 Published
Benchmark	M2.7 Published
SWE-Pro	56.2% (matches GPT-5.3-Codex)
SWE Multilingual	76.5%
Multi-SWE-Bench	52.7%
MLE Bench Lite	66.6% medal rate (22 ML competitions)
VIBE-Pro	55.6% (near Opus 4.6)
TerminalBench 2	57.0%
NL2Repo	39.8%
GDPval-AA ELO	1495 (highest open-weight)
Toolathon	46.3% accuracy
MM Claw (skill compliance)	97% across 40+ skills
MM Claw (end-to-end)	62.7% (near Sonnet 4.6)

SLURPY — best of both

SLURPY's merge schedule preserves M2.5's deep reasoning character in the early-to-mid layers (where the two models barely differ) while absorbing M2.7's agentic improvements in the late layers (where M2.7's training signal concentrates). The result is a model that carries both parents' strengths without the training cost of either.

Merge method

Per-tensor empirical SLERP — each of the 48,239 mergeable weight tensors gets its own interpolation ratio t(k) derived from the measured cosine similarity between M2.5 and M2.7 on that specific tensor:

text

delta(k)      = 1 - cos(M2.5_k, M2.7_k)delta_norm(k) = clip(delta(k) / delta_p99, 0, 1)t(k)          = 0.50 + 0.35 * delta_norm(k)

Tensors that barely changed (cos ~ 1.0): t ~ 0.50 — neutral midpoint, preserving both parents
Tensors that changed the most (layer 61 MoE experts): t = 0.85 — absorbing M2.7's concentrated training signal
FP8 weights: dequantized to BF16 before SLERP, re-quantized with fresh block-wise scales
No scale_inv pass-through: forensics confirmed 0% bit-identical scales between parents — all 47,864 FP8 scale tensors are recomputed, not copied

Forensic highlights

99.18% of tensors sit in a tight cosine cluster around 0.9946 — most weights barely moved between M2.5 and M2.7
Layer 61 MoE experts {76, 74, 61, 30, 43, 138, 226, 126, 58, 159} have deltas 2-5x baseline — this is where M2.7's RL training signal concentrates
lm_head.weight (cos=0.9905, rel_l2=0.139) carries M2.7's vocabulary-level improvements

Architecture

Identical to MiniMax-M2.5 / M2.7 — weight merge only, no architecture changes:

Model type: minimax_m2 / MiniMaxM2ForCausalLM
Parameters: 228.7B total, ~10B active (MoE)
Layers: 62
Hidden size: 3072
MoE: 256 experts, top-8, sigmoid routing + learned bias
Attention: 48 query / 8 KV heads (GQA 6:1), head_dim=128
Quantization: FP8 (float8_e4m3fn), block size [128, 128]
Vocab: 200,064 tokens
Context: up to 196,608 tokens
Thinking: Interleaved <think>...</think> (always-on)
trust_remote_code=True required

Serving with vLLM

Recommended command (8x H100 80GB):

bash

SAFETENSORS_FAST_GPU=1 vllm serve \    Ex0bit/MiniMax-SLURPY --trust-remote-code \    --enable-expert-parallel --tensor-parallel-size 8 \    --enable-auto-tool-choice --tool-call-parser minimax_m2 \    --reasoning-parser minimax_m2_append_think \    --enforce-eager

For 4x GPU (no expert parallel):

bash

SAFETENSORS_FAST_GPU=1 vllm serve \    Ex0bit/MiniMax-SLURPY --trust-remote-code \    --tensor-parallel-size 4 \    --enable-auto-tool-choice --tool-call-parser minimax_m2 \    --reasoning-parser minimax_m2_append_think

If you encounter CUDA memory errors, add:

bash

--compilation-config '{"cudagraph_mode": "PIECEWISE"}'

Recommended sampling parameters

Table with columns: Parameter, Value
Parameter	Value
temperature	1.0
top_p	0.95
top_k	40

Important: preserve thinking in conversation history

MiniMax-M2 uses interleaved thinking. The model outputs <think>...</think> blocks during generation. You must pass these back verbatim in conversation history. Removing them degrades performance.

Tool calling

Same format as MiniMax-M2.7. Tool calls use <minimax:tool_call> / </minimax:tool_call> XML wrappers:

xml

<minimax:tool_call><invoke name="get_weather"><parameter name="city">San Francisco</parameter></invoke></minimax:tool_call>

Enable with --enable-auto-tool-choice --tool-call-parser minimax_m2 in vLLM.

Using with Transformers

python

from transformers import AutoModelForCausalLM, AutoTokenizerimport torch model = AutoModelForCausalLM.from_pretrained(    "Ex0bit/MiniMax-SLURPY",    trust_remote_code=True,    torch_dtype="auto",    device_map="auto",)tokenizer = AutoTokenizer.from_pretrained(    "Ex0bit/MiniMax-SLURPY",    trust_remote_code=True,) messages = [{"role": "user", "content": "Write a Python function that reverses a linked list."}]input_ids = tokenizer.apply_chat_template(    messages, add_generation_prompt=True, return_tensors="pt").to(model.device) with torch.no_grad():    output = model.generate(        input_ids,        max_new_tokens=2048,        do_sample=True,        temperature=1.0,        top_p=0.95,        top_k=40,    ) print(tokenizer.decode(output[0, input_ids.shape[1]:], skip_special_tokens=True))

Config notes

use_mtp is set to False in config.json (MTP tensors don't exist in the checkpoint)
quantization_config is preserved — native FP8
Chat template and tokenizer are sourced from M2.7

Files

43 safetensors shards (~5 GB each, 214.3 GB total)
Native FP8 (float8_e4m3fn) with block-wise [128, 128] scale factors
chat_template.jinja — M2.7's chat template with tool calling support
modeling_minimax_m2.py / configuration_minimax_m2.py — custom model code

License

Modified MIT — same as MiniMax-M2.5. See LICENSE for full text.

The only modification to the standard MIT license: if the Software (or any derivative works) is used for commercial products or services with more than 100 million monthly active users or more than $30M annual recurring revenue, you must prominently display "MiniMax M2" on the user interface.

Citation

text

@misc{minimax-slurpy-2026,  title={MiniMax-SLURPY: Per-tensor empirical SLERP merge of MiniMax-M2.5 and M2.7},  author={Ex0bit},  year={2026},  url={https://huggingface.co/Ex0bit/MiniMax-SLURPY}}

Acknowledgments

MiniMax for the M2.5 and M2.7 base models
Merge infrastructure adapted from the PRISM abliteration pipeline

Available on FriendliAI

Dedicated Endpoints

Run this model inference on single tenant GPU with unmatched speed and reliability at scale.

Learn more

Model Details

Model Provider

Ex0bit

Model Tree

Base

MiniMaxAI/MiniMax-M2.7

Base

MiniMaxAI/MiniMax-M2.5

Merged

this model

Input Modalities

Text

Output Modalities

Text

Supported Functionality

Dedicated Endpoints

Explore FriendliAI today

Get started Talk to an engineer

README

License: modified-mit

What SLURPY inherits

From M2.5 — the architect

M2.5 is the foundation-builder: strong on greenfield engineering, deep reasoning, and research-grade benchmarks.

Table with columns: Benchmark, M2.5 Published
Benchmark	M2.5 Published
SWE-Bench Verified	80.2%
BrowseComp (with context mgmt)	76.3%
Multi-SWE-Bench	51.3%
AIME 2025	86.3
GPQA Diamond	85.2
SciCode	44.4
IFBench	70.0
HLE (w/o tools)	19.4
GDPval-MM (office work)	59.0% avg win rate

From M2.7 — the operator

M2.7 is the execution specialist: RL-tuned for multi-step tool use, terminal ops, agentic scaffolding, and production-grade software engineering.

Table with columns: Benchmark, M2.7 Published
Benchmark	M2.7 Published
SWE-Pro	56.2% (matches GPT-5.3-Codex)
SWE Multilingual	76.5%
Multi-SWE-Bench	52.7%
MLE Bench Lite	66.6% medal rate (22 ML competitions)
VIBE-Pro	55.6% (near Opus 4.6)
TerminalBench 2	57.0%
NL2Repo	39.8%
GDPval-AA ELO	1495 (highest open-weight)
Toolathon	46.3% accuracy
MM Claw (skill compliance)	97% across 40+ skills
MM Claw (end-to-end)	62.7% (near Sonnet 4.6)

SLURPY — best of both

Merge method

text

delta(k)      = 1 - cos(M2.5_k, M2.7_k)delta_norm(k) = clip(delta(k) / delta_p99, 0, 1)t(k)          = 0.50 + 0.35 * delta_norm(k)

delta(k)      = 1 - cos(M2.5_k, M2.7_k)delta_norm(k) = clip(delta(k) / delta_p99, 0, 1)t(k)          = 0.50 + 0.35 * delta_norm(k)

Tensors that barely changed (cos ~ 1.0): t ~ 0.50 — neutral midpoint, preserving both parents
Tensors that changed the most (layer 61 MoE experts): t = 0.85 — absorbing M2.7's concentrated training signal
FP8 weights: dequantized to BF16 before SLERP, re-quantized with fresh block-wise scales
No scale_inv pass-through: forensics confirmed 0% bit-identical scales between parents — all 47,864 FP8 scale tensors are recomputed, not copied

Forensic highlights

99.18% of tensors sit in a tight cosine cluster around 0.9946 — most weights barely moved between M2.5 and M2.7
Layer 61 MoE experts {76, 74, 61, 30, 43, 138, 226, 126, 58, 159} have deltas 2-5x baseline — this is where M2.7's RL training signal concentrates
lm_head.weight (cos=0.9905, rel_l2=0.139) carries M2.7's vocabulary-level improvements

Architecture

Identical to MiniMax-M2.5 / M2.7 — weight merge only, no architecture changes:

Model type: minimax_m2 / MiniMaxM2ForCausalLM
Parameters: 228.7B total, ~10B active (MoE)
Layers: 62
Hidden size: 3072
MoE: 256 experts, top-8, sigmoid routing + learned bias
Attention: 48 query / 8 KV heads (GQA 6:1), head_dim=128
Quantization: FP8 (float8_e4m3fn), block size [128, 128]
Vocab: 200,064 tokens
Context: up to 196,608 tokens
Thinking: Interleaved <think>...</think> (always-on)
trust_remote_code=True required

Serving with vLLM

Recommended command (8x H100 80GB):

bash

SAFETENSORS_FAST_GPU=1 vllm serve \    Ex0bit/MiniMax-SLURPY --trust-remote-code \    --enable-expert-parallel --tensor-parallel-size 8 \    --enable-auto-tool-choice --tool-call-parser minimax_m2 \    --reasoning-parser minimax_m2_append_think \    --enforce-eager

SAFETENSORS_FAST_GPU=1 vllm serve \    Ex0bit/MiniMax-SLURPY --trust-remote-code \    --enable-expert-parallel --tensor-parallel-size 8 \    --enable-auto-tool-choice --tool-call-parser minimax_m2 \    --reasoning-parser minimax_m2_append_think \    --enforce-eager

For 4x GPU (no expert parallel):

bash

SAFETENSORS_FAST_GPU=1 vllm serve \    Ex0bit/MiniMax-SLURPY --trust-remote-code \    --tensor-parallel-size 4 \    --enable-auto-tool-choice --tool-call-parser minimax_m2 \    --reasoning-parser minimax_m2_append_think

SAFETENSORS_FAST_GPU=1 vllm serve \    Ex0bit/MiniMax-SLURPY --trust-remote-code \    --tensor-parallel-size 4 \    --enable-auto-tool-choice --tool-call-parser minimax_m2 \    --reasoning-parser minimax_m2_append_think

If you encounter CUDA memory errors, add:

bash

--compilation-config '{"cudagraph_mode": "PIECEWISE"}'

Recommended sampling parameters

Table with columns: Parameter, Value
Parameter	Value
temperature	1.0
top_p	0.95
top_k	40

Important: preserve thinking in conversation history

Tool calling

Same format as MiniMax-M2.7. Tool calls use <minimax:tool_call> / </minimax:tool_call> XML wrappers:

xml

<minimax:tool_call><invoke name="get_weather"><parameter name="city">San Francisco</parameter></invoke></minimax:tool_call>

<minimax:tool_call><invoke name="get_weather"><parameter name="city">San Francisco</parameter></invoke></minimax:tool_call>

Enable with --enable-auto-tool-choice --tool-call-parser minimax_m2 in vLLM.

Using with Transformers

python

from transformers import AutoModelForCausalLM, AutoTokenizerimport torch model = AutoModelForCausalLM.from_pretrained(    "Ex0bit/MiniMax-SLURPY",    trust_remote_code=True,    torch_dtype="auto",    device_map="auto",)tokenizer = AutoTokenizer.from_pretrained(    "Ex0bit/MiniMax-SLURPY",    trust_remote_code=True,) messages = [{"role": "user", "content": "Write a Python function that reverses a linked list."}]input_ids = tokenizer.apply_chat_template(    messages, add_generation_prompt=True, return_tensors="pt").to(model.device) with torch.no_grad():    output = model.generate(        input_ids,        max_new_tokens=2048,        do_sample=True,        temperature=1.0,        top_p=0.95,        top_k=40,    ) print(tokenizer.decode(output[0, input_ids.shape[1]:], skip_special_tokens=True))

from transformers import AutoModelForCausalLM, AutoTokenizerimport torch model = AutoModelForCausalLM.from_pretrained(    "Ex0bit/MiniMax-SLURPY",    trust_remote_code=True,    torch_dtype="auto",    device_map="auto",)tokenizer = AutoTokenizer.from_pretrained(    "Ex0bit/MiniMax-SLURPY",    trust_remote_code=True,) messages = [{"role": "user", "content": "Write a Python function that reverses a linked list."}]input_ids = tokenizer.apply_chat_template(    messages, add_generation_prompt=True, return_tensors="pt").to(model.device) with torch.no_grad():    output = model.generate(        input_ids,        max_new_tokens=2048,        do_sample=True,        temperature=1.0,        top_p=0.95,        top_k=40,    ) print(tokenizer.decode(output[0, input_ids.shape[1]:], skip_special_tokens=True))

Config notes

use_mtp is set to False in config.json (MTP tensors don't exist in the checkpoint)
quantization_config is preserved — native FP8
Chat template and tokenizer are sourced from M2.7

Files

43 safetensors shards (~5 GB each, 214.3 GB total)
Native FP8 (float8_e4m3fn) with block-wise [128, 128] scale factors
chat_template.jinja — M2.7's chat template with tool calling support
modeling_minimax_m2.py / configuration_minimax_m2.py — custom model code

License

Modified MIT — same as MiniMax-M2.5. See LICENSE for full text.

Citation

text

@misc{minimax-slurpy-2026,  title={MiniMax-SLURPY: Per-tensor empirical SLERP merge of MiniMax-M2.5 and M2.7},  author={Ex0bit},  year={2026},  url={https://huggingface.co/Ex0bit/MiniMax-SLURPY}}

@misc{minimax-slurpy-2026,  title={MiniMax-SLURPY: Per-tensor empirical SLERP merge of MiniMax-M2.5 and M2.7},  author={Ex0bit},  year={2026},  url={https://huggingface.co/Ex0bit/MiniMax-SLURPY}}

Acknowledgments

MiniMax for the M2.5 and M2.7 base models
Merge infrastructure adapted from the PRISM abliteration pipeline