Up to date
google Gemma Terms of Use out 3 years ago
8.54B parameters, text in, text out, 8K tokens of context.
Verified 2026-10-25 Rank #662 in LLMs · score 72.9 method v1 Sources 3 Downloads 30d 51.2k · #380 of 619 tracked
Picked by rule from 2 tracked pieces: the lab's own post, the most shared, a field report, the latest.
The 1 pick above, plus these 1. 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: 10 quantized builds and 21 Spaces.
Top Spaces running it: ysharma/CodeGemma ♥ 90 · Justinrune/LLaMA-Factory ♥ 12 · NiansuhAI/Main ♥ 10
Weights only, an estimate: leave room for the KV cache. Method
About 5.1 GB at INT4, 10 GB at FP8, 21 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 (2024-03-21), amber triangles are 2 dated articles (each links to it), grey ticks are repo commits.
26 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.
| Benchmark | Value | Setting / notes | Reported by | Date |
|---|---|---|---|---|
| author-reported BC HE C# | 26.7 | 7B-IT | model author | 2026-10-07 |
| author-reported BC HE C++ | 42.2 | 7B-IT | model author | 2026-10-07 |
| author-reported BC HE Go | 28.6 | 7B-IT | model author | 2026-10-07 |
| author-reported BC HE Java | 48.4 | 7B-IT | model author | 2026-10-07 |
| author-reported BC HE JavaScript | 46 | 7B-IT | model author | 2026-10-07 |
| author-reported BC HE Kotlin | 51.6 | 7B-IT | model author | 2026-10-07 |
| author-reported BC HE Python | 48.4 | 7B-IT | model author | 2026-10-07 |
| author-reported BC HE Rust | 36 | 7B-IT | model author | 2026-10-07 |
| author-reported HumanEval · Write code with agents | 56.1 | 7B-IT | model author | 2026-10-07 |
| author-reported HumanEval · Write code with agents | 68.25 | 7B-IT | model author | 2026-10-07 |
| author-reported HumanEval · Write code with agents | 20.05 | 7B-IT | model author | 2026-10-07 |
| author-reported MBPP · Write code with agents | 54.2 | 7B-IT | model author | 2026-10-07 |
| author-reported MBPP · Write code with agents | 56.7 | 7B-IT | model author | 2026-10-07 |
| author-reported MBPP · Write code with agents | 41.2 | 7B-IT | model author | 2026-10-07 |
| author-reported MBPP · Write code with agents | 46.2 | 7B-IT | model author | 2026-10-07 |
| author-reported MBPP · Write code with agents | 57.3 | 7B-IT | model author | 2026-10-07 |
| author-reported MBPP · Write code with agents | 61.4 | 7B-IT | model author | 2026-10-07 |
| author-reported MBPP · Write code with agents | 59.9 | 7B-IT | model author | 2026-10-07 |
| author-reported MBPP · Write code with agents | 62 | 7B-IT | model author | 2026-10-07 |
| author-reported MBPP · Write code with agents | 53.5 | 7B-IT | model author | 2026-10-07 |
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 | GemmaForCausalLM |
| Model type | gemma |
| Dtype | bfloat16 |
| Params by dtype | BF16: 8,537,680,896 |
| Size band | M |
| MoE | no |
| Context length | 8,192 (config.json max_position_embeddings) |
| Gated | manual |
| License | Gemma Terms of Use (gemma) |
| Hub snapshot | 2026-10-25 |
vLLM: supported · GemmaForCausalLM · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · GemmaForCausalLM · 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 codegemma-7b-it: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 21 GB · Single GPU (<24 GB) |
|---|---|
| FP8 (1 byte/param) | 10 GB · Single GPU (<24 GB) |
| INT4 (0.5 bytes/param) | 5.1 GB · Single GPU (<24 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.
quantized: 10finetune: 10adapter: 6
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| MaziyarPanahi/codegemma-7b-it-GGUF | quantized | GGUF | — | — | — | MaziyarPanahi | 0 | 10.2k |
| mradermacher/codegemma-7b-it-i1-GGUF | quantized | GGUF | — | — | gemma | mradermacher | 0 | 8.3k |
| second-state/CodeGemma-7b-it-GGUF | quantized | GGUF | — | — | gemma | second-state | 2 | 6.3k |
| mradermacher/codegemma-7b-it-GGUF | quantized | GGUF | — | — | gemma | mradermacher | 1 | 3.1k |
| alphaduriendur/codegemma-7b-it-Q4_K_M-GGUF | quantized | GGUF | — | — | gemma | alphaduriendur | 0 | 551 |
| anthonym21/pinescript-v5-instructions-merged | finetune | — | 8.54B | M | apache-2.0 | anthonym21 | 2 | 307 |
| sheldonrobinson/codegemma-7b-it-Q8_0-GGUF | quantized | Q8_0 | — | — | gemma | sheldonrobinson | 0 | 374 |
| bms22/codegemma-7b-it-Q4_K_M-GGUF | quantized | GGUF | — | — | gemma | bms22 | 0 | 372 |
| ernie-research/APPS-Gemma-7B-MA-PPO-Fixed10 | finetune | — | 8.54B | M | mit | ernie-research | 1 | 254 |
| sheldonrobinson/codegemma-7b-it-Q4_K_M-GGUF | quantized | GGUF | — | — | gemma | sheldonrobinson | 0 | 299 |
| lapfed255/codegemma-7b-it-Q4_K_M-GGUF | quantized | GGUF | — | — | gemma | lapfed255 | 0 | 273 |
| scthornton/codegemma-7b-securecode | adapter | — | — | — | gemma | scthornton | 0 | 235 |
| zeetroid/code-bench-CodeGemma-7BIT-cg-nv9n_it_zs | adapter | — | — | — | gemma | zeetroid | 0 | 182 |
| llmware/codegemma-7b-it-ov | quantized | — | — | — | gemma | llmware | 1 | 99 |
| zeetroid/code-bench-CodeGemma-7BIT-cg-nv9n_it_fs | adapter | — | — | — | gemma | zeetroid | 0 | 119 |
| lurf21/codegemma-7b-it-NEP | finetune | — | 9.32B | L | apache-2.0 | lurf21 | 0 | 112 |
| kawasumi/Tema_Q-R2.0-Code | finetune | — | 8.54B | M | gemma | kawasumi | 0 | 96 |
| ShethArihant/PSC-2_codegemma-7b-it_sft_2-epochs | adapter | — | — | — | — | ShethArihant | 0 | 70 |
| anthonym21/pinescript-v5-instructions | finetune | — | — | — | — | anthonym21 | 1 | 0 |
| litert-community/codegemma-7b-it-int4-litertlm | finetune | int4 | — | — | gemma | litert-community | 1 | 0 |
| KakashiH/BashExplainer_Gemma | adapter | — | — | — | apache-2.0 | KakashiH | 0 | 40 |
| LDES777/tap_ft_gemma | adapter | — | — | — | — | LDES777 | 0 | 21 |
| gaianet/CodeGemma-7b-it-GGUF | finetune | GGUF | — | — | gemma | gaianet | 0 | 0 |
| athenadai/codegemma-finetuned-qlora-rank4 | finetune | — | — | — | — | athenadai | 0 | 0 |
| ShethArihant/codegemma-7b-it_sft_2-epochs | finetune | — | — | — | — | ShethArihant | 0 | 0 |
| cpxavier2/codegemma-7b-it-int4-litertlm | finetune | int4 | — | — | gemma | cpxavier2 | 0 | 0 |
| Date | Commit | SHA |
|---|---|---|
| 2024-08-07 | Upload fixed tokenizer files (#11) | 078cdc5 |
| 2024-06-27 | Update link to terms (#10) | eeb9fb1 |
| 2024-04-16 | Fix tokenizer (#7) | 9b1d923 |
| 2024-04-11 | Update README.md (#8) | 14a1c06 |
| 2024-04-09 | Update README.md | 6b02e58 |
| 2024-04-09 | Tokenizer Fixes (#5) | 2a467a0 |
| 2024-04-09 | Squashing commit | dd73e23 |
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.
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:42 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:13 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.
