- Size
- 39.8B 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
- 21 Oct 2026the date the repo was created
- Context length
- 256k tokensfrom the model's config
- Kind
- Base model (no base model declared on the Hub)FlintForCausalLM
- On Hugging Face
- RecursiveSI/flint62k downloads in the last 30 days, 1.4k 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: 0.25 GB, worked out from the model's config (RecursiveSI/flint), 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
unsloth/flint-GGUF (converted by unsloth), 352k downloads. Sizes are the real files on the Hub, all parts added up.
| Format | Size | Memory needed | Get it |
|---|---|---|---|
| UD-IQ1_M | 10.6 GBfile size | 11.3 GB10.6 + 0.25 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/flint-GGUF:UD-IQ1_MOllama ollama run hf.co/unsloth/flint-GGUF:UD-IQ1_MDownload hf download unsloth/flint-GGUF --include "flint-UD-IQ1_M.gguf" --local-dir flint-GGUF |
| UD-IQ2_XXS | 11.0 GBfile size | 11.8 GB11.0 + 0.25 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/flint-GGUF:UD-IQ2_XXSOllama ollama run hf.co/unsloth/flint-GGUF:UD-IQ2_XXSDownload hf download unsloth/flint-GGUF --include "flint-UD-IQ2_XXS.gguf" --local-dir flint-GGUF |
| UD-Q2_K_XL | 12.6 GBfile size | 13.4 GB12.6 + 0.25 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/flint-GGUF:UD-Q2_K_XLOllama ollama run hf.co/unsloth/flint-GGUF:UD-Q2_K_XLDownload hf download unsloth/flint-GGUF --include "flint-UD-Q2_K_XL.gguf" --local-dir flint-GGUF |
| UD-IQ3_XXS | 13.5 GBfile size | 14.3 GB13.5 + 0.25 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/flint-GGUF:UD-IQ3_XXSOllama ollama run hf.co/unsloth/flint-GGUF:UD-IQ3_XXSDownload hf download unsloth/flint-GGUF --include "flint-UD-IQ3_XXS.gguf" --local-dir flint-GGUF |
| UD-Q3_K_XL | 17.2 GBfile size | 18.0 GB17.2 + 0.25 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/flint-GGUF:UD-Q3_K_XLOllama ollama run hf.co/unsloth/flint-GGUF:UD-Q3_K_XLDownload hf download unsloth/flint-GGUF --include "flint-UD-Q3_K_XL.gguf" --local-dir flint-GGUF |
| UD-Q4_K_XL | 23.3 GBfile size | 24.1 GB23.3 + 0.25 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/flint-GGUF:UD-Q4_K_XLOllama ollama run hf.co/unsloth/flint-GGUF:UD-Q4_K_XLDownload hf download unsloth/flint-GGUF --include "flint-UD-Q4_K_XL.gguf" --local-dir flint-GGUF |
| UD-Q5_K_XL | 27.9 GBfile size | 28.6 GB27.9 + 0.25 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/flint-GGUF:UD-Q5_K_XLOllama ollama run hf.co/unsloth/flint-GGUF:UD-Q5_K_XLDownload hf download unsloth/flint-GGUF --include "flint-UD-Q5_K_XL.gguf" --local-dir flint-GGUF |
| UD-Q6_K_XL | 33.8 GBfile size | 34.6 GB33.8 + 0.25 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/flint-GGUF:UD-Q6_K_XLOllama ollama run hf.co/unsloth/flint-GGUF:UD-Q6_K_XLDownload hf download unsloth/flint-GGUF --include "flint-UD-Q6_K_XL.gguf" --local-dir flint-GGUF |
| Q8_0 | 39.4 GBfile size | 40.1 GB39.4 + 0.25 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/flint-GGUF:Q8_0Ollama ollama run hf.co/unsloth/flint-GGUF:Q8_0Download hf download unsloth/flint-GGUF --include "flint-Q8_0.gguf" --local-dir flint-GGUF |
| BF16 | 74.1 GBfile size, 3 parts | 74.9 GB74.1 + 0.25 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/flint-GGUF:BF16Download hf download unsloth/flint-GGUF --include "BF16/flint-BF16-*-of-00003.gguf" --local-dir flint-GGUF |
MLX for Apple silicon
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| MLX mxfp4RESMP-DEV/Flint-40B-A2.5B-MLX-GPTQ-MXFP4 | 20.6 GBfile size | 21.3 GB20.6 + 0.25 + 0.50 | Download page |
| MLX 4bitpgmharikrishnan/Flint-MLX-Serve-4bit | 20.9 GBfile size | 21.6 GB20.9 + 0.25 + 0.50 | Download page |
Original weights size from the safetensors header, for vLLM, SGLang, Transformers
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| BF16original weights | 74.1 GBfile size | 74.9 GB74.1 + 0.25 + 0.50 |
hf download RecursiveSI/flint --local-dir flintOther conversions (14; sizes not read)
- trymirai/Flint-S Other, 972 downloads
- infosave/flint-cmf Other, 577 downloads
- FutureMLS-Lab/flint-nvfp4-coldgptq GPTQ, nvfp4, 516 downloads
- mattbucci/Flint-AWQ AWQ, 452 downloads
- cyankiwi/flint-AWQ-INT4 AWQ, int4, 368 downloads
- 0xBakeer/TandemLLM-Flint-EXL3-6.0 EXL3, 225 downloads
- HawkBearPig/flint-FP8-BF16 FP8, 216 downloads
- JRamirez-UAB/Flint-40B-A2.5B-EXL3-3.88bpw-24-48GB EXL3, 211 downloads
- HawkBearPig/flint-FP8-NVFP4 NVFP4, fp8, 172 downloads
- 0xBakeer/TandemLLM-Flint-EXL3-4.0 EXL3, 139 downloads
- HawkBearPig/flint-NVFP4-FP8 NVFP4, 138 downloads
- yamz-labs/Flint-40B-A2.5B-EXL3-Yamz EXL3, 126 downloads
- urakozz/flint-40B-A2.5B-W4A16-g64-AutoRound-GPTQ GPTQ, w4a16, 118 downloads
- Xylolabs/flint-Tenstorrent-Tensix-BFP4 Other, 107 downloads
The Hub tags 100 or more 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.
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.