AxForge
Sign in Get started

Model reference · open weights

starcoderbase

LLMs bigcode Text gen 2 builds Its own licence terms 3k dl/mo

starcoderbase is an open-weight language model from bigcode. starcoderbase (BF16) weighs 63.3 GB; the smallest configuration that runs it is H100 80 GB.

  • StarCoderBase is a 15.5B parameter text-generation model developed by BigCode for code generation.
  • It was trained on 1 trillion tokens from 80+ programming languages and features an 8192 token context window.
  • The model is licensed under the BigCode OpenRAIL-M license.

Summary of the bigcode/starcoderbase model card, 2026-10-01

Hugging Face ↗GitHub ↗Paper ↗

What it is

Released bybigcode
Released2023-05-03
VRAM63.3 GB for the weights

What it runs on

Memory and cards for starcoderbase (BF16)

63.3 GBweights, file size
762 MBruntime overhead, at least

How much memory each request adds isn't estimated yet for this architecture. The weights need at least the cards below, plus room for the context.

CardWeights alone
RTX 3060 12 GB … L40S 48 GB
6 smaller cards		does not fit
A100 80 GBfits
H100 80 GBtight
RTX PRO 6000 Blackwell 96 GBfits
DGX Spark (GB10) 128 GB unifiedfits
H200 141 GBfits
B200 180 GBfits

Builds

Sizes, precisions & builds

BuildParamsPrecisionWeightsSmallest setup
starcoderbase (above) ↗ —BF16 63.3 GBH100 80 GB
starcoderbase-1b ↗ 1.1BFP32 2.3 GBRTX 3060 12 GB

How it works

How language models work

Your prompttext / messagesTransformerattention over tokensNext-token loopgenerate + streamResponsetext · tool callsA language model reads your tokens and predicts the next one, again and again, streaming the reply back.

Benchmarks

Reported results

As published on the model card: the maker's own numbers, not measured by AxForge.

TaskDatasetMetricScore
text-generationHumanEvalpass@10.304
text-generationMBPPpass@10.490
text-generationDS-1000 (Overall Completion)pass@10.238
text-generationMultiPL-HumanEval (C++)pass@10.306
text-generationMultiPL-HumanEval (C#)pass@10.206
text-generationMultiPL-HumanEval (D)pass@10.100
text-generationMultiPL-HumanEval (Go)pass@10.215
text-generationMultiPL-HumanEval (Java)pass@10.285
text-generationMultiPL-HumanEval (Julia)pass@10.211
text-generationMultiPL-HumanEval (JavaScript)pass@10.317
text-generationMultiPL-HumanEval (Lua)pass@10.266
text-generationMultiPL-HumanEval (PHP)pass@10.268
text-generationMultiPL-HumanEval (Perl)pass@10.163
text-generationMultiPL-HumanEval (Python)pass@10.303
text-generationMultiPL-HumanEval (R)pass@10.102
text-generationMultiPL-HumanEval (Ruby)pass@10.172
text-generationMultiPL-HumanEval (Racket)pass@10.118
text-generationMultiPL-HumanEval (Rust)pass@10.245
text-generationMultiPL-HumanEval (Scala)pass@10.288
text-generationMultiPL-HumanEval (Bash)pass@10.110
text-generationMultiPL-HumanEval (Swift)pass@10.167
text-generationMultiPL-HumanEval (TypeScript)pass@10.322
© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms