Model
Model explorer

Mistral Large 24.11

OPEN
Mistral AI · Mistral Large family · released Nov 18, 2024

Nov 2024 refresh of Large 2 with improved long-context handling, function calling and system prompt support.

ReasoningCodingVisionFunction callingTool useAgentic
653.6
Elo · rank #288
Parameters
123B
Active params
123B (dense)
Context
128K tokens
Architecture
Dense Transformer
License
Mistral Research License (non-commercial; commercial license required for production)
Languages
—
API price (in/out)
No hosted API
Modalities
text
Benchmark results
Bar shows position within the tracked field; marker = field best
BIG-Bench HardReasoning52.7%#94
best: ERNIE 4.5 300B-A47B · 94.3%
GPQA DiamondReasoning24.9%#285
best: GPT-6 Astra · 96.0%
IFEvalReasoning84.0%#71
best: Gemma 4 26B A4B · 98.5%
MMLU-ProKnowledge67.9%#99
best: Claude Fable 5 · 91.5%
Run it locally
VRAM @ Q4
74 GB
VRAM @ FP16
246 GB
Fits on (Q4)
M3 Max 128GBM3 Ultra 512GBA100 80GBH100 80GBH200 141GBB200 192GB
123B model exceeds a single RTX 4090's 24GB VRAM even at Q4; requires multi-GPU or CPU offload.
Quantizations
GGUF Q4 · AWQ · MLX
Fine-tune it
Research only
QLoRA83.6 GB1× H200 141GB
LoRA262.0 GB2× H200 141GB
Full fine-tune1974.2 GBbeyond 8× B200
QLoRA SFT on ~10k samples ≈ $63.61 (1× H200 141GB)
Plan a fine-tune →
Mistral Large family
Elo progression across releases
Mistral LargeFeb 2024661.8Mistral Large 2 (24.07)Jul 20241025.1Mistral Large 24.11Nov 2024653.6Mistral Large 3Dec 20251283.1
Resources
HFmistralai/Mistral-Large-Instruct-2411DOCSdocs.mistral.ai/models/mistral-large-2-1-24-11
API price weights · each benchmark row carries its own source badge (see methodology)