Model
Model explorer

Stable LM 2 12B

OPEN
Stability AI · StableLM 2 family · released Apr 8, 2024

Mid-size multilingual flagship of StableLM 2; DPO-tuned chat variant adds tool-use and function-calling support.

ReasoningCodingVisionFunction callingTool useAgentic
440.3
Elo · rank #321
Parameters
12.1B
Active params
12.1B (dense)
Context
4K tokens
Architecture
Dense decoder-only transformer (40 layers, GQA 32 query/8 KV heads, RoPE, parallel attn+FFN)
License
Stability AI Community License (non-commercial by default; commercial use requires Stability AI membership)
Languages
7+
API price (in/out)
No hosted API
Modalities
text
Benchmark results
Bar shows position within the tracked field; marker = field best
ARC-ChallengeReasoning65.0%#64
best: Llama 3.1 405B · 96.9%
GSM8KMath57.7%#119
best: Llama 3.1 405B · 96.8%
HellaSwagReasoning86.1%#30
best: Claude 3 Opus · 95.4%
MMLUKnowledge61.1%#200
best: OpenAI o3 · 92.9%
MT-BenchHuman preference8.15#43
best: Hunyuan-Large (A52B) · 9.4
TruthfulQAKnowledge62.0%#21
best: Phi-3.5-MoE (16x3.8B, 6.6B active) · 77.5%
WinoGrandeReasoning78.8%#42
best: PaLM 2 · 90.9%
Run it locally
VRAM @ Q4
8 GB
VRAM @ FP16
24.5 GB
Fits on (Q4)
RTX 3060 12GBRTX 4070 Ti 16GBRTX 3090 24GBRTX 4090 24GBRTX 5090 32GBM4 Pro 48GBM3 Max 128GBM3 Ultra 512GBA100 80GBH100 80GBH200 141GBB200 192GB
Throughput data unavailable.
Quantizations
GGUF
StableLM 2 family
Elo progression across releases
Stable LM 2 1.6BJan 2024-229.9Stable LM 2 12BApr 2024440.3
Resources
HFstabilityai/stablelm-2-12bGHStability-AI/StableLM
API price weights · each benchmark row carries its own source badge (see methodology)