ibm-granite · fine-tune of ibm-granite/granite-3.3-8b-base
granite-3.3-8b-instruct
Runs on your machine
Choose a machine
We then show the largest format that fits, from real file sizes where we have them.
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-3.3-8b-instruct), at 16-bit precision. Choose a machine above to see a verdict for each format. Method
weights context cache runtime
GGUF for llama.cpp, Ollama, LM Studio and others
ibm-granite/granite-3.3-8b-instruct-GGUF (the maker's own conversion), 8.6k downloads. Sizes are the real files on the Hub, all parts added up. Showing the common levels.
| Format | Size | Memory needed | Memory | Get it |
|---|---|---|---|---|
| Q4_K_M | 4.6 GBfile size | 6.4 GB4.6 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-3.3-8b-instruct-GGUF:Q4_K_MOllama ollama run hf.co/ibm-granite/granite-3.3-8b-instruct-GGUF:Q4_K_MDownload hf download ibm-granite/granite-3.3-8b-instruct-GGUF --include "granite-3.3-8b-instruct-Q4_K_M.gguf" --local-dir granite-3.3-8b-instruct-GGUF | |
| Q5_K_M | 5.4 GBfile size | 7.1 GB5.4 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-3.3-8b-instruct-GGUF:Q5_K_MOllama ollama run hf.co/ibm-granite/granite-3.3-8b-instruct-GGUF:Q5_K_MDownload hf download ibm-granite/granite-3.3-8b-instruct-GGUF --include "granite-3.3-8b-instruct-Q5_K_M.gguf" --local-dir granite-3.3-8b-instruct-GGUF | |
| Q6_K | 6.2 GBfile size | 8.0 GB6.2 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-3.3-8b-instruct-GGUF:Q6_KOllama ollama run hf.co/ibm-granite/granite-3.3-8b-instruct-GGUF:Q6_KDownload hf download ibm-granite/granite-3.3-8b-instruct-GGUF --include "granite-3.3-8b-instruct-Q6_K.gguf" --local-dir granite-3.3-8b-instruct-GGUF | |
| Q8_0 | 8.1 GBfile size | 9.8 GB8.1 + 1.3 + 0.50 | Commandsllama.cpp llama-server -hf ibm-granite/granite-3.3-8b-instruct-GGUF:Q8_0Ollama ollama run hf.co/ibm-granite/granite-3.3-8b-instruct-GGUF:Q8_0Download hf download ibm-granite/granite-3.3-8b-instruct-GGUF --include "granite-3.3-8b-instruct-Q8_0.gguf" --local-dir granite-3.3-8b-instruct-GGUF |
MLX for Apple silicon
| Format | Size | Memory needed | Memory | Repo |
|---|---|---|---|---|
| MLX 4bitmlx-community/granite-3.3-8b-instruct-4bit | 4.3 GBfile size | 6.0 GB4.3 + 1.3 + 0.50 | Download page | |
| MLX 8bitmlx-community/granite-3.3-8b-instruct-8bit | 8.1 GBfile size | 9.8 GB8.1 + 1.3 + 0.50 | Download page |
Original weights size from the safetensors header, for vLLM, SGLang, Transformers
| Format | Size | Memory needed | Memory | Repo |
|---|---|---|---|---|
| BF16original weights | 15.2 GBfile size | 17.0 GB15.2 + 1.3 + 0.50 |
hf download ibm-granite/granite-3.3-8b-instruct --local-dir granite-3.3-8b-instructOther conversions (8; sizes not read)
- ibm-granite/granite-3.3-8b-instruct-FP8 FP8, by the maker, 1.7k downloads
- drawais/Granite-3.3-8B-Instruct-AWQ-INT4 AWQ, int4, 55 downloads
- TevunahAi/Granite-3.3-8B-Instruct-GPTQ GPTQ, 18 downloads
- stelterlab/granite-3.3-8b-instruct-FP8-Dynamic FP8, 16 downloads
- cgus/granite-3.3-8b-instruct-exl2 EXL2, 10 downloads
- keisuke-miyako/granite-3.3-8b-instruct-onnx-int4 Other, int4, 10 downloads
- owentruong/granite-3.3-8b-instruct-EXL2 EXL2, 0 downloads
- Prince-1/Granite-3.3-8B-Instruct-Onnx Other, 0 downloads
The Hub tags 35 quantised repos as conversions of this model; we read the most downloaded and show the maker's own first.
Run it by API
No provider on Hugging Face Inference Providers serves this model (as of 19 Nov 2026, 13:58 UTC). Other hosts may; we only list what the router reports.
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.
Fine-tunes of this model
- ibm-granite/granite-3.3-8b-math-prm-v28.2B2.5kUnreviewed
- ibm-granite/granite-3.3-8b-lora-math-prm0Unreviewed
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.