Model
Model explorer

SantaCoder 1.1B

OPEN
BigCode · SantaCoder family · released Dec 22, 2022

BigCode's first model: 1.1B multilingual (Python/Java/JS) code LLM with FIM infilling, proof-of-concept for The Stack.

ReasoningCodingVisionFunction callingTool useAgentic
-153.5
Elo · unrated
Parameters
1.1B
Active params
1.1B (dense)
Context
2K tokens
Architecture
GPT-2-style decoder with multi-query attention (GPTBigCode) + fill-in-the-middle
License
BigCode OpenRAIL-M
Languages
3+
API price (in/out)
No hosted API
Modalities
text
Benchmark results
Bar shows position within the tracked field; marker = field best
HumanEval FIMCoding44.0%#7
best: Codestral 22B · 91.6%
HumanEvalCoding18.0%#171
best: Claude Opus 4.5 · 99.4%
MBPPCoding35.0%#92
best: Llama-3.3-Nemotron-Super-49B v1 (Reasoning On) · 91.3%
Run it locally
VRAM @ Q4
0.8 GB
VRAM @ FP16
2.3 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
—
Fine-tune it
Conditional / custom
QLoRA2.7 GB1× RTX 3060 12GB
LoRA4.3 GB1× RTX 3060 12GB
Full fine-tune19.6 GB1× RTX 3090 24GB
QLoRA SFT on ~10k samples ≈ $0.64 (1× RTX 3060 12GB)
Plan a fine-tune →
SantaCoder family
Elo progression across releases
SantaCoder 1.1BDec 2022-153.5
Resources
HFbigcode/santacoderDOCSbigcode-project.org/docs/about/models
API price weights · each benchmark row carries its own source badge (see methodology)