AxForge
Sign in Get started

Model reference · open weights

starcoder

LLMs bigcode Text gen 1 build Its own licence terms 6k dl/mo

starcoder is an open-weight language model from bigcode. starcoder (FP32) weighs 31.6 GB; the smallest configuration that runs it is L40S 48 GB.

Request Rent a GPU

For self-hosting, compare this build's requirements below with the GPU you plan to rent.

  • StarCoder is a 15.8B parameter text-generation model developed by BigCode for code generation across 80+ programming languages.
  • It features an 8192 token context window and was trained on 1 trillion tokens using the Fill-in-the-Middle objective.
  • The model is released under the BigCode OpenRAIL-M license.

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

Hugging Face ↗GitHub ↗Paper ↗

What it is

Released bybigcode
Released2023-04-24
Parameters15.8B
VRAM31.6 GB for the weights

What it runs on

Memory and cards for starcoder (FP32)

31.6 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 … RTX 5090 32 GB
5 smaller cards		does not fit
L40S 48 GBfits
A100 80 GBfits
H100 80 GBfits
RTX PRO 6000 Blackwell 96 GBfits
DGX Spark (GB10) 128 GB unifiedfits
H200 141 GBfits
B200 180 GBfits

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-generationHumanEval (Prompted)pass@10.408
text-generationHumanEvalpass@10.336
text-generationMBPPpass@10.527
text-generationDS-1000 (Overall Completion)pass@10.260
text-generationMultiPL-HumanEval (C++)pass@10.316
text-generationMultiPL-HumanEval (C#)pass@10.210
text-generationMultiPL-HumanEval (D)pass@10.136
text-generationMultiPL-HumanEval (Go)pass@10.176
text-generationMultiPL-HumanEval (Java)pass@10.302
text-generationMultiPL-HumanEval (Julia)pass@10.230
text-generationMultiPL-HumanEval (JavaScript)pass@10.308
text-generationMultiPL-HumanEval (Lua)pass@10.239
text-generationMultiPL-HumanEval (PHP)pass@10.261
text-generationMultiPL-HumanEval (Perl)pass@10.173
text-generationMultiPL-HumanEval (Python)pass@10.336
text-generationMultiPL-HumanEval (R)pass@10.155
text-generationMultiPL-HumanEval (Ruby)pass@10.012
text-generationMultiPL-HumanEval (Racket)pass@10.001
text-generationMultiPL-HumanEval (Rust)pass@10.218
text-generationMultiPL-HumanEval (Scala)pass@10.276
text-generationMultiPL-HumanEval (Bash)pass@10.105
text-generationMultiPL-HumanEval (Swift)pass@10.227
text-generationMultiPL-HumanEval (TypeScript)pass@10.323
© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms