- Size
- 8.8B parameterscounted from the safetensors header on the Hub
- Licence
- unknown, Apache License 2.0Our reading, check the licence: You can use, change and redistribute the model, including commercially. Keep the licence and any notices, and say which files you changed; you do not have to publish your changes. It includes a patent licence that ends for anyone who sues over patents in the work.
- On the Hub since
- 7 Aug 2026the date the repo was created
- Context length
- 128k tokensfrom the model's config
- Fine-tune
- of ibm-granite/granite-4.1-8b-baseGraniteForCausalLM
- On Hugging Face
- ibm-granite/granite-4.2-8b58k downloads in the last 30 days, 128 likes
Scores
No independent scores for this model in Epoch AI's data. A score is only shown here when Epoch's name for a model matches this repo with certainty; an uncertain match is left out rather than guessed.
Claimed by the maker Claimed
The model card has no claimed scores in a machine-readable form (the card's model-index), so none are shown. The maker may give scores in the card text: read the card.
Other views, which we link to and do not copy: look this model up on Artificial Analysis and the Arena leaderboard.
Run it yourself
Memory needed = weights + context cache + 0.50 GB for the runtime. Context cache at 8k tokens: 1.3 GB, worked out from the model's config (ibm-granite/granite-4.2-8b), at 16-bit precision. Choose a machine above to see a verdict for each format. Method
GGUF for llama.cpp, Ollama, LM Studio and others
ibm-granite/granite-4.2-8b-GGUF (the maker's own conversion), 281k downloads. Sizes are the real files on the Hub, all parts added up.
| Format | Size | Memory needed | Get it |
|---|---|---|---|
| Q3_K_S | 3.7 GBfile size | 5.4 GB3.7 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-4.2-8b-GGUF:Q3_K_SOllama ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:Q3_K_SDownload hf download ibm-granite/granite-4.2-8b-GGUF --include "granite-4.2-8b-Q3_K_S.gguf" --local-dir granite-4.2-8b-GGUF |
| Q3_K_M | 4.0 GBfile size | 5.8 GB4.0 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-4.2-8b-GGUF:Q3_K_MOllama ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:Q3_K_MDownload hf download ibm-granite/granite-4.2-8b-GGUF --include "granite-4.2-8b-Q3_K_M.gguf" --local-dir granite-4.2-8b-GGUF |
| Q3_K_L | 4.4 GBfile size | 6.1 GB4.4 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-4.2-8b-GGUF:Q3_K_LOllama ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:Q3_K_LDownload hf download ibm-granite/granite-4.2-8b-GGUF --include "granite-4.2-8b-Q3_K_L.gguf" --local-dir granite-4.2-8b-GGUF |
| Q4_0 | 4.7 GBfile size | 6.5 GB4.7 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-4.2-8b-GGUF:Q4_0Ollama ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:Q4_0Download hf download ibm-granite/granite-4.2-8b-GGUF --include "granite-4.2-8b-Q4_0.gguf" --local-dir granite-4.2-8b-GGUF |
| Q4_K_S | 4.7 GBfile size | 6.5 GB4.7 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-4.2-8b-GGUF:Q4_K_SOllama ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:Q4_K_SDownload hf download ibm-granite/granite-4.2-8b-GGUF --include "granite-4.2-8b-Q4_K_S.gguf" --local-dir granite-4.2-8b-GGUF |
| Q4_K_M | 5.0 GBfile size | 6.7 GB5.0 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-4.2-8b-GGUF:Q4_K_MOllama ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:Q4_K_MDownload hf download ibm-granite/granite-4.2-8b-GGUF --include "granite-4.2-8b-Q4_K_M.gguf" --local-dir granite-4.2-8b-GGUF |
| Q4_1 | 5.2 GBfile size | 6.9 GB5.2 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-4.2-8b-GGUF:Q4_1Ollama ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:Q4_1Download hf download ibm-granite/granite-4.2-8b-GGUF --include "granite-4.2-8b-Q4_1.gguf" --local-dir granite-4.2-8b-GGUF |
| Q5_0 | 5.7 GBfile size | 7.4 GB5.7 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-4.2-8b-GGUF:Q5_0Ollama ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:Q5_0Download hf download ibm-granite/granite-4.2-8b-GGUF --include "granite-4.2-8b-Q5_0.gguf" --local-dir granite-4.2-8b-GGUF |
| Q5_K_S | 5.7 GBfile size | 7.4 GB5.7 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-4.2-8b-GGUF:Q5_K_SOllama ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:Q5_K_SDownload hf download ibm-granite/granite-4.2-8b-GGUF --include "granite-4.2-8b-Q5_K_S.gguf" --local-dir granite-4.2-8b-GGUF |
| Q5_K_M | 5.8 GBfile size | 7.6 GB5.8 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-4.2-8b-GGUF:Q5_K_MOllama ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:Q5_K_MDownload hf download ibm-granite/granite-4.2-8b-GGUF --include "granite-4.2-8b-Q5_K_M.gguf" --local-dir granite-4.2-8b-GGUF |
| Q5_1 | 6.2 GBfile size | 7.9 GB6.2 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-4.2-8b-GGUF:Q5_1Ollama ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:Q5_1Download hf download ibm-granite/granite-4.2-8b-GGUF --include "granite-4.2-8b-Q5_1.gguf" --local-dir granite-4.2-8b-GGUF |
| Q6_K | 6.7 GBfile size | 8.5 GB6.7 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-4.2-8b-GGUF:Q6_KOllama ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:Q6_KDownload hf download ibm-granite/granite-4.2-8b-GGUF --include "granite-4.2-8b-Q6_K.gguf" --local-dir granite-4.2-8b-GGUF |
| Q8_0 | 8.7 GBfile size | 10.5 GB8.7 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-4.2-8b-GGUF:Q8_0Ollama ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:Q8_0Download hf download ibm-granite/granite-4.2-8b-GGUF --include "granite-4.2-8b-Q8_0.gguf" --local-dir granite-4.2-8b-GGUF |
| BF16 | 16.4 GBfile size | 18.1 GB16.4 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-4.2-8b-GGUF:BF16Ollama ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:BF16Download hf download ibm-granite/granite-4.2-8b-GGUF --include "granite-4.2-8b-bf16.gguf" --local-dir granite-4.2-8b-GGUF |
MLX for Apple silicon
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| MLXibm-granite/granite-4.2-8b-q4-mlx | 4.6 GBfile size | 6.4 GB4.6 + 1.3 + 0.50 | Download page |
| MLXibm-granite/granite-4.2-8b-q8-mlx | 8.7 GBfile size | 10.4 GB8.7 + 1.3 + 0.50 | Download page |
Original weights size from the safetensors header, for vLLM, SGLang, Transformers
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| BF16original weights | 16.4 GBfile size | 18.1 GB16.4 + 1.3 + 0.50 |
hf download ibm-granite/granite-4.2-8b --local-dir granite-4.2-8bOther conversions (6; sizes not read)
- ibm-granite/granite-4.2-8b-fp8 FP8, by the maker, 5.2k downloads
- ibm-granite/granite-4.2-8b-nvfp4 NVFP4, by the maker, 2.5k downloads
- ibm-granite/granite-4.2-8b-mxfp4 Other, mxfp4, by the maker, 2.4k downloads
- Xylolabs/granite-4.2-8b-Tenstorrent-Tensix-BFP8 Other, 89 downloads
- Xylolabs/granite-4.2-8b-Tenstorrent-Tensix-BFP4 Other, 77 downloads
- Jeethu/granite-4.2-8b-PARO Other, 73 downloads
The Hub tags 66 quantised repos as conversions of this model; we read the most downloaded and show the maker's own first.
Run it by API
2 providers serve this model through the Hugging Face router. Prices are US dollars per million tokens. Precision is what the provider reports; a lower precision than the original weights can change answers. Live first, then cheapest output.
| Provider | In | Out | Context | Precision | Tokens/s | First token | Uptime 24h | Tools | JSON |
|---|---|---|---|---|---|---|---|---|---|
| CoreWeave | $0.06 | $0.15 | 128k | bf16 | 91 | 0.2 s | 99.2% | yes | yes |
| DeepInfra | $0.06 | $0.25 | 128k | bf16 | 90 | 0.5 s | 100.0% | yes | no |
Through the router, with a Hugging Face token (the router bills your Hugging Face account). Change the name after the colon to pick another provider.
curl https://router.huggingface.co/v1/chat/completions \
-H "Authorization: Bearer $HF_TOKEN" -H "Content-Type: application/json" \
-d '{"model": "ibm-granite/granite-4.2-8b:coreweave", "messages": [{"role": "user", "content": "Hello"}]}'Not built yet: one key for every open model, with BlockRun, and our own tested-quality score per provider. Until then, precision and measured speed are the best guide to a provider here.
Planned for this page Not built yet
- Try it in private chat
- A link to Solana AI's confidential chat for the models it serves. This model is not listed as served there.
- Fine-tune it
- Fine-tuning a few base models from this page and publishing the result here.
- Creators and earnings
- Pages for people who publish fine-tunes and adapters of this model, and what they earn.
- Fund and bounties
- Bounties for new open models, paid from a public fund. How the fund will work.
- Contribute compute
- How to give compute to open model work, with Akash.
Model data from the Hugging Face Hub 19 Nov 2026, 13:43 UTC. The repo. Found a mistake on this page? Tell us.