Up to date
bigcode BigCode OpenRAIL-M out 4 years ago
15.8B parameters, text in, text out.
Verified 2026-10-25 Rank #604 in LLMs · score 74 method v1 Sources 3 Downloads 30d 2.7k · #572 of 619 tracked
Picked by rule from 12 tracked pieces: the lab's own post, the most shared, a field report, the latest.
StarCoder and StarCoderBase are 15.5B parameter models with 8K context length and infilling capabilities.
BigCode released StarChat, a fine-tuned version of the 16B parameter StarCoder model trained on one trillion tokens from 80+ programming languages.
llama.cpp now supports StarCoder model architectures, including 1B, 3B, 7B, and 15B variants.
The 4 picks above, plus these 8. Top is the radar's score: discussion, source authority and the lab's own posts, fading with age.
Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.
No API provider serves it, so its downloads are people running it themselves: 7 quantized builds and 100 Spaces.
Top Spaces running it: bigcode/bigcode-playground ♥ 414 · KBaba7/Quant ♥ 51 · bigcode/search ♥ 40 · Yeyito/llm_contamination_detector ♥ 27
Weights only, an estimate: leave room for the KV cache. Method
About 9.5 GB at INT4, 19 GB at FP8, 38 GB at BF16: weights only, so leave room for the KV cache.
Estimate params × bytes per param × 1.2 · how we estimate
Weights, every dated article, and repo commits on one rail.
Teal is the weights (2023-04-24), amber triangles are 10 dated articles (each links to it).
77 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Everything else, folded: the full tables, every source.
Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.
Listed in LLMs
| Pipeline tag | text-generation |
|---|---|
| Library | transformers |
| Architectures | GPTBigCodeForCausalLM |
| Model type | gpt_bigcode |
| Params by dtype | F32: 15,819,446,272 |
| Size band | L |
| MoE | no |
| Gated | auto |
| Papers | arXiv:1911.02150, arXiv:2205.14135, arXiv:2207.14255, arXiv:2305.06161 |
| License | BigCode OpenRAIL-M (bigcode-openrail-m) |
| Hub snapshot | 2026-10-25 |
vLLM: supported · GPTBigCodeForCausalLM · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · GPTBigCodeForCausalLM · 773cb938fe9484a8f7e7e46fb1b1fe96412c59ed
vLLM registry (7050ef881ff0835f6890a47db1859a24c65481d6) · SGLang registry (773cb938fe9484a8f7e7e46fb1b1fe96412c59ed) · Matched by architecture class name against each project's registry on main; 'supported' means the architecture is registered, not that this checkpoint was tested.
Weights-only memory for starcoder: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 38 GB · Single high-end GPU (<80 GB) |
|---|---|
| FP8 (1 byte/param) | 19 GB · Single GPU (<24 GB) |
| INT4 (0.5 bytes/param) | 9.5 GB · Single GPU (<24 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.
quantized: 7finetune: 33adapter: 37
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| mradermacher/starcoder-i1-GGUF | quantized | GGUF | — | — | bigcode-openrail-m | mradermacher | 1 | 16.7k |
| mradermacher/starcoder-GGUF | quantized | GGUF | — | — | bigcode-openrail-m | mradermacher | 2 | 12.5k |
| DevQuasar-9/bigcode.starcoder-GGUF | quantized | GGUF | — | — | — | DevQuasar-9 | 0 | 3.3k |
| jlpan/SteloCoder | finetune | — | — | — | bigcode-openrail-m | jlpan | 1 | 2.3k |
| smangrul/peft-lora-starcoder15B-v2-personal-copilot-A100-40GB-colab | adapter | — | — | — | bigcode-openrail-m | smangrul | 5 | 722 |
| osukhoroslov-hw/starcoder-Q5_K_M-GGUF | quantized | GGUF | — | — | bigcode-openrail-m | osukhoroslov-hw | 0 | 719 |
| AdnanRiaz107/starcoder-Q4_K_M-GGUF | quantized | GGUF | — | — | bigcode-openrail-m | AdnanRiaz107 | 0 | 361 |
| jlpan/starcoder-tune-cpp2py-program1 | adapter | — | — | — | bigcode-openrail-m | jlpan | 0 | 235 |
| jlpan/starcoder-csharp2py-program2 | adapter | — | — | — | bigcode-openrail-m | jlpan | 1 | 154 |
| szkelo/starcoderbase-3b-q8_0 | quantized | q8_0 | — | — | bigcode-openrail-m | szkelo | 0 | 180 |
| smangrul/peft-lora-starcoder15B-personal-copilot-A100-40GB-colab | adapter | — | — | — | bigcode-openrail-m | smangrul | 0 | 158 |
| KarthiAru/peft-lora-starcoder-personal-copilot-A100-40GB-colab | adapter | — | — | — | bigcode-openrail-m | KarthiAru | 0 | 155 |
| jlpan/starcoder-tune-cpp2py-program3 | adapter | — | — | — | bigcode-openrail-m | jlpan | 0 | 149 |
| Esperanto/starcoder-kvc-fp16-onnx | quantized | -fp16 | — | — | — | Esperanto | 0 | 134 |
| jlpan/starcoder-php2py-snippet1 | adapter | — | — | — | bigcode-openrail-m | jlpan | 0 | 129 |
| jlpan/starcoder-php2py-snippet2 | adapter | — | — | — | bigcode-openrail-m | jlpan | 0 | 120 |
| taylorbobaylor/peft-lora-starcoder15B-v2-personal-copilot-A100-40GB-colab | adapter | — | — | — | bigcode-openrail-m | taylorbobaylor | 0 | 115 |
| jlpan/starcoder-c2py-testprogram | adapter | — | — | — | bigcode-openrail-m | jlpan | 0 | 110 |
| stillerman/magic-starcoder | adapter | — | — | — | bigcode-openrail-m | stillerman | 0 | 110 |
| stillerman/starcoder-mathematica | adapter | — | — | — | bigcode-openrail-m | stillerman | 1 | 57 |
| jlpan/starcoder-js2py-snippet1 | adapter | — | — | — | bigcode-openrail-m | jlpan | 0 | 106 |
| jlpan/starcoder-java2py-program2 | adapter | — | — | — | bigcode-openrail-m | jlpan | 0 | 100 |
| jlpan/starcoder-js2py-snippet3 | adapter | — | — | — | bigcode-openrail-m | jlpan | 0 | 96 |
| terryyz/starcoder-codecot | adapter | — | — | — | — | terryyz | 0 | 65 |
| jlpan/starcoder-tune-cpp2py-program2 | adapter | — | — | — | bigcode-openrail-m | jlpan | 0 | 62 |
| jlpan/starcoder-js2py-program2 | adapter | — | — | — | bigcode-openrail-m | jlpan | 0 | 56 |
| htahir1/peft-lora-zencoder15B-personal-copilot | adapter | — | — | — | bigcode-openrail-m | htahir1 | 0 | 53 |
| jlpan/starcoder-finetuned-code2code | finetune | — | — | — | bigcode-openrail-m | jlpan | 1 | 0 |
| jlpan/starcoder-csharp2py-program1 | finetune | — | — | — | bigcode-openrail-m | jlpan | 1 | 0 |
| jlpan/starcoder-csharp2py-snippet1 | finetune | — | — | — | bigcode-openrail-m | jlpan | 1 | 0 |
| jlpan/starcoder-csharp2py-snippet2 | finetune | — | — | — | bigcode-openrail-m | jlpan | 1 | 0 |
| jlpan/starcoder-php2py-program1 | adapter | — | — | — | bigcode-openrail-m | jlpan | 0 | 44 |
| rreit/starcoder-StaproCoder | adapter | — | — | — | bigcode-openrail-m | rreit | 0 | 44 |
| Diyer/peft-lora-starcoder15B-v2-personal-copilot-A100-40GB-colab | adapter | — | — | — | — | Diyer | 0 | 37 |
| update0909/peft-lora-starcoder15B-v2-personal-copilot-A100-40GB-colab-zyx | adapter | — | — | — | bigcode-openrail-m | update0909 | 0 | 34 |
| Myrax3000/peft-lora-starcoder15B-personal-copilot-T4-colab | adapter | — | — | — | — | Myrax3000 | 0 | 33 |
| update0909/peft-lora-starcoder15B-personal-copilot-A100-40GB-colab | adapter | — | — | — | bigcode-openrail-m | update0909 | 0 | 33 |
| update0909/peft-lora-starcoder15B-personal-copilot-A100-40GB-colab-zyx | adapter | — | — | — | bigcode-openrail-m | update0909 | 0 | 33 |
| sallywww/StarCoder_oneStep_FuzzTargets | adapter | — | — | — | — | sallywww | 0 | 33 |
| titanpark/peft-lora-starcoder15B-v2-personal-copilot-A100-40GB-colab | adapter | — | — | — | — | titanpark | 0 | 30 |
Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency. Summaries are generated from a stored copy of the article text; every sentence cites a span of that text, and nothing is shown that failed the citation check.
arXiv links found in the model card and its tags. "Peer-reviewed" is shown only with evidence: a published venue in OpenAlex (measured) or the authors' own journal reference on arXiv (reported).
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:39 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:10 UTC · Last verified 2026-10-25
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.
