LEADBOARD - ADMET, Kinase and Toxicity Prediction Benchmark
π―
31
Benchmark for drug prediction tools - ADMET, kinase, tox
99
cutechicken
AI & ML interests
None yet
Recent Activity
liked a Space about 9 hours ago
FINAL-Bench/leadboard reacted to SeaWolf-AI's post with π₯ about 9 hours ago
We opened a benchmark for drug property prediction tools. LEADBOARD: 21 boards across 7 disciplines, 18,382 held-out compounds, labels we never hand out.
Two numbers we hit while building it are the reason it exists.
First. Split the hERG cardiotoxicity data at random and you get AUROC 0.818. Split it by first-report year instead and you get 0.606. Same molecules, same fingerprints, same learner, same hyperparameters. The only thing that changed was where the line went, and the score moved 0.211. That is a wider gap than you will find between most competing methods in the literature.
Second. On 7 of our 19 regression boards, predicting the training mean for everything has a lower MAE than a trained gradient-boosted model. hERG is one of them, 0.599 against 0.589. The trained model loses.
So every board publishes its homework before anyone submits. Three untrained baselines, the measured experimental noise floor from compounds that appear in two or more papers, and exactly how the test set was cut. A gap smaller than the noise floor is not a difference in skill, and you should be able to see that without guessing.
Entering is simple. Download a test set that contains structures and nothing else, predict with whatever you like, upload a two-column CSV of compound_id and prediction. Trained model, physics engine, LLM, rule of thumb. We do not care what is inside. We measure the output.
Post: https://huggingface.co/blog/FINAL-Bench/leadboard-drug
Leaderboard: https://huggingface.co/spaces/FINAL-Bench/leadboard reacted to SeaWolf-AI's post with π₯ 1 day ago
π Open Materials Challenge, Season 1 β Solid-State Battery Electrolytes
A solid-state battery replaces the liquid electrolyte of a lithium-ion cell with a solid. It does not catch fire, it lasts longer, and it can hold more. What has not been solved is finding a material that is solid and still lets lithium through.
Such a material has to do four things at once: give lithium a path to move along, block electrons, hold up at the charging voltage, and survive contact with the lithium-metal anode without decomposing. Plenty of materials manage three. Very few manage all four.
This challenge looks for candidates, together. You submit one composition β for example Li3YCl6. We score it computationally and place it on the board. There is no prize.
Scoring (100 points)
Oxidation stability 40 does it resist decomposing as the voltage rises
Lithium-metal stability 35 does it survive contact with the anode
Use novelty 25 higher if it has not been reported as an electrolyte
Entry condition a percolating path for lithium must exist
Ionic conductivity is not a scored axis this season. Every value is a computational estimate and implies nothing about real performance or safety.
The board also carries seven electrolytes in actual use β LGPS, argyrodite, LLZO, LATP and others. They are scored but hold no rank. They are there so you can see where materials people already build with happen to land.
Compositions are private by default. Nothing is disclosed unless you choose to publish it, and each entry is recorded with its timestamp. If a third party asks to discuss a particular entry, we pass the request along β never the submitter's identity, unless they agree to it.
Season 1 runs 2026-08-21 to 11-30. A participation guide and a set of prompts are included.
π https://huggingface.co/spaces/FINAL-Bench/open-materials-challenge
Our sister challengeπ§ͺhttps://huggingface.co/spaces/FINAL-Bench/open-discovery-challengeOrganizations
liked a
Space about 9 hours ago
reacted to SeaWolf-AI's post with π₯ about 9 hours ago
Post
838
We opened a benchmark for drug property prediction tools. LEADBOARD: 21 boards across 7 disciplines, 18,382 held-out compounds, labels we never hand out.
Two numbers we hit while building it are the reason it exists.
First. Split the hERG cardiotoxicity data at random and you get AUROC 0.818. Split it by first-report year instead and you get 0.606. Same molecules, same fingerprints, same learner, same hyperparameters. The only thing that changed was where the line went, and the score moved 0.211. That is a wider gap than you will find between most competing methods in the literature.
Second. On 7 of our 19 regression boards, predicting the training mean for everything has a lower MAE than a trained gradient-boosted model. hERG is one of them, 0.599 against 0.589. The trained model loses.
So every board publishes its homework before anyone submits. Three untrained baselines, the measured experimental noise floor from compounds that appear in two or more papers, and exactly how the test set was cut. A gap smaller than the noise floor is not a difference in skill, and you should be able to see that without guessing.
Entering is simple. Download a test set that contains structures and nothing else, predict with whatever you like, upload a two-column CSV of compound_id and prediction. Trained model, physics engine, LLM, rule of thumb. We do not care what is inside. We measure the output.
Post: https://huggingface.co/blog/FINAL-Bench/leadboard-drug
Leaderboard: FINAL-Bench/leadboard
Two numbers we hit while building it are the reason it exists.
First. Split the hERG cardiotoxicity data at random and you get AUROC 0.818. Split it by first-report year instead and you get 0.606. Same molecules, same fingerprints, same learner, same hyperparameters. The only thing that changed was where the line went, and the score moved 0.211. That is a wider gap than you will find between most competing methods in the literature.
Second. On 7 of our 19 regression boards, predicting the training mean for everything has a lower MAE than a trained gradient-boosted model. hERG is one of them, 0.599 against 0.589. The trained model loses.
So every board publishes its homework before anyone submits. Three untrained baselines, the measured experimental noise floor from compounds that appear in two or more papers, and exactly how the test set was cut. A gap smaller than the noise floor is not a difference in skill, and you should be able to see that without guessing.
Entering is simple. Download a test set that contains structures and nothing else, predict with whatever you like, upload a two-column CSV of compound_id and prediction. Trained model, physics engine, LLM, rule of thumb. We do not care what is inside. We measure the output.
Post: https://huggingface.co/blog/FINAL-Bench/leadboard-drug
Leaderboard: FINAL-Bench/leadboard
- 2 replies
reacted to SeaWolf-AI's post with π₯ 1 day ago
Post
2930
π Open Materials Challenge, Season 1 β Solid-State Battery Electrolytes
A solid-state battery replaces the liquid electrolyte of a lithium-ion cell with a solid. It does not catch fire, it lasts longer, and it can hold more. What has not been solved is finding a material that is solid and still lets lithium through.
Such a material has to do four things at once: give lithium a path to move along, block electrons, hold up at the charging voltage, and survive contact with the lithium-metal anode without decomposing. Plenty of materials manage three. Very few manage all four.
This challenge looks for candidates, together. You submit one composition β for example Li3YCl6. We score it computationally and place it on the board. There is no prize.
Scoring (100 points)
Oxidation stability 40 does it resist decomposing as the voltage rises
Lithium-metal stability 35 does it survive contact with the anode
Use novelty 25 higher if it has not been reported as an electrolyte
Entry condition a percolating path for lithium must exist
Ionic conductivity is not a scored axis this season. Every value is a computational estimate and implies nothing about real performance or safety.
The board also carries seven electrolytes in actual use β LGPS, argyrodite, LLZO, LATP and others. They are scored but hold no rank. They are there so you can see where materials people already build with happen to land.
Compositions are private by default. Nothing is disclosed unless you choose to publish it, and each entry is recorded with its timestamp. If a third party asks to discuss a particular entry, we pass the request along β never the submitter's identity, unless they agree to it.
Season 1 runs 2026-08-21 to 11-30. A participation guide and a set of prompts are included.
π FINAL-Bench/open-materials-challenge
Our sister challengeπ§ͺhttps://huggingface.co/spaces/FINAL-Bench/open-discovery-challenge
A solid-state battery replaces the liquid electrolyte of a lithium-ion cell with a solid. It does not catch fire, it lasts longer, and it can hold more. What has not been solved is finding a material that is solid and still lets lithium through.
Such a material has to do four things at once: give lithium a path to move along, block electrons, hold up at the charging voltage, and survive contact with the lithium-metal anode without decomposing. Plenty of materials manage three. Very few manage all four.
This challenge looks for candidates, together. You submit one composition β for example Li3YCl6. We score it computationally and place it on the board. There is no prize.
Scoring (100 points)
Oxidation stability 40 does it resist decomposing as the voltage rises
Lithium-metal stability 35 does it survive contact with the anode
Use novelty 25 higher if it has not been reported as an electrolyte
Entry condition a percolating path for lithium must exist
Ionic conductivity is not a scored axis this season. Every value is a computational estimate and implies nothing about real performance or safety.
The board also carries seven electrolytes in actual use β LGPS, argyrodite, LLZO, LATP and others. They are scored but hold no rank. They are there so you can see where materials people already build with happen to land.
Compositions are private by default. Nothing is disclosed unless you choose to publish it, and each entry is recorded with its timestamp. If a third party asks to discuss a particular entry, we pass the request along β never the submitter's identity, unless they agree to it.
Season 1 runs 2026-08-21 to 11-30. A participation guide and a set of prompts are included.
π FINAL-Bench/open-materials-challenge
Our sister challengeπ§ͺhttps://huggingface.co/spaces/FINAL-Bench/open-discovery-challenge
liked a
Space 1 day ago
reacted to SeaWolf-AI's post with π₯β€οΈ 3 days ago
Post
3915
3,631 candidate molecules arrived in five days, from 83 accounts β roughly 700 a day. Far more than we expected. Thank you.
Yesterday we opened the third season and 224 arrived within a day: Chagas disease.
Why this disease
Around 6 million people live with it, mostly in Latin America (WHO). Many carry it for decades without knowing, while the heart is slowly damaged. There are two drugs and both date from the 1960s, hard enough to tolerate that many patients cannot finish the two-month course.
Sixty years without a new drug is not only a scientific problem. Most patients live where development costs cannot be recovered, which is why WHO calls this a neglected tropical disease.
But the cost of proposing a candidate and filtering it has changed. So it seemed worth asking whether work nobody funds could be done by many people sharing it out.
The problem this season
The target is CYP51, the enzyme T. cruzi uses to build its membrane sterols. Block it and the parasite cannot survive. The difficulty is that we carry the same enzyme.
Scoring: binding 30 Β· selectivity 30 Β· ADMET 15 Β· whole-cell 10 Β· novelty 10 Β· synthesis 5
Selectivity carries 30 points because nobody has solved it. Among the approved azoles on the board as reference compounds, some score 0 on selectivity β not a scorer fault, but the measurement.
Taking part
Design with any model, submit a SMILES, scored within minutes. Five ready-to-paste prompts per season, and the full rubric is published. Your molecule stays yours; private submission is the default.
Prizes β 4,000 USD across three seasons
Malaria 30 Sep Β· 1,000 | Tuberculosis 31 Oct Β· 2,000 | Chagas 30 Nov Β· 1,000
We know this does not cover the time you spend. It is a way of saying the work had worth.
FINAL-Bench/open-discovery-challenge
Yesterday we opened the third season and 224 arrived within a day: Chagas disease.
Why this disease
Around 6 million people live with it, mostly in Latin America (WHO). Many carry it for decades without knowing, while the heart is slowly damaged. There are two drugs and both date from the 1960s, hard enough to tolerate that many patients cannot finish the two-month course.
Sixty years without a new drug is not only a scientific problem. Most patients live where development costs cannot be recovered, which is why WHO calls this a neglected tropical disease.
But the cost of proposing a candidate and filtering it has changed. So it seemed worth asking whether work nobody funds could be done by many people sharing it out.
The problem this season
The target is CYP51, the enzyme T. cruzi uses to build its membrane sterols. Block it and the parasite cannot survive. The difficulty is that we carry the same enzyme.
Scoring: binding 30 Β· selectivity 30 Β· ADMET 15 Β· whole-cell 10 Β· novelty 10 Β· synthesis 5
Selectivity carries 30 points because nobody has solved it. Among the approved azoles on the board as reference compounds, some score 0 on selectivity β not a scorer fault, but the measurement.
Taking part
Design with any model, submit a SMILES, scored within minutes. Five ready-to-paste prompts per season, and the full rubric is published. Your molecule stays yours; private submission is the default.
Prizes β 4,000 USD across three seasons
Malaria 30 Sep Β· 1,000 | Tuberculosis 31 Oct Β· 2,000 | Chagas 30 Nov Β· 1,000
We know this does not cover the time you spend. It is a way of saying the work had worth.
FINAL-Bench/open-discovery-challenge
upvoted a collection 7 days ago
liked a
Space 8 days ago
upvoted an article 8 days ago
Article
Who Tells You Whether the Molecule Your AI Just Designed Is Any Good?
FINAL-Bench
β’ β’ 11 reacted to SeaWolf-AI's post with π 8 days ago
Post
4270
𧬠Your AI can design a malaria drug candidate. Can it tell you whether it's any good?
Open Discovery Challenge #1 β Malaria is live. Design a molecule with any model β OpenAI, Claude, Gemini, Qwen, KIMI, DeepSeek, open weights, or by hand β submit it as SMILES, and it's scored in minutes on whole-cell activity, target binding, selectivity over the human enzyme, ADMET, novelty and synthesisability.
You can check the scoring instead of trusting it. Approved drugs sit on the same leaderboard as the entries: DSM265, a clinical-stage antimalarial, scores 50.9. Teriflunomide β approved, but it hits the human enzyme β scores 2.8. Caffeine scores 1.8. If the clinical candidate lands on top and coffee lands at the bottom, the scorer discriminates.
We caught 14 defects before opening β conventional toxicity cutoffs rejected all three approved antimalarials and coffee. All written up, along with the rule we now hold everything to: a gate that rejects an approved drug is a broken gate.
Your molecule stays yours. No patent interest, nothing into our pipeline. You choose whether it's published β and publishing can cost you patentability, so we say so.
USD 1,000 to the top entry when Season #1 closes 30 September 2026 β not payment for your tokens, but a way of saying the work had worth.
Malaria killed ~597,000 people in 2023, three quarters of them children under five. Not for want of chemistry β for want of a market.
No chemistry needed: the guide ships five prompts you can paste straight into your model, and the full rubric is published.
π https://huggingface.co/blog/FINAL-Bench/open-discovery-challenge
π FINAL-Bench/open-discovery-challenge
Computational assessments of candidates β not measurements, not claims of efficacy.
Open Discovery Challenge #1 β Malaria is live. Design a molecule with any model β OpenAI, Claude, Gemini, Qwen, KIMI, DeepSeek, open weights, or by hand β submit it as SMILES, and it's scored in minutes on whole-cell activity, target binding, selectivity over the human enzyme, ADMET, novelty and synthesisability.
You can check the scoring instead of trusting it. Approved drugs sit on the same leaderboard as the entries: DSM265, a clinical-stage antimalarial, scores 50.9. Teriflunomide β approved, but it hits the human enzyme β scores 2.8. Caffeine scores 1.8. If the clinical candidate lands on top and coffee lands at the bottom, the scorer discriminates.
We caught 14 defects before opening β conventional toxicity cutoffs rejected all three approved antimalarials and coffee. All written up, along with the rule we now hold everything to: a gate that rejects an approved drug is a broken gate.
Your molecule stays yours. No patent interest, nothing into our pipeline. You choose whether it's published β and publishing can cost you patentability, so we say so.
USD 1,000 to the top entry when Season #1 closes 30 September 2026 β not payment for your tokens, but a way of saying the work had worth.
Malaria killed ~597,000 people in 2023, three quarters of them children under five. Not for want of chemistry β for want of a market.
No chemistry needed: the guide ships five prompts you can paste straight into your model, and the full rubric is published.
π https://huggingface.co/blog/FINAL-Bench/open-discovery-challenge
π FINAL-Bench/open-discovery-challenge
Computational assessments of candidates β not measurements, not claims of efficacy.
- 12 replies
reacted to SeaWolf-AI's post with π 9 days ago
Post
3293
AX-Ray: Safety Diagnostics for AI/AX Models
AI models can no longer be evaluated only by capability scores. As models move into public services, enterprise workflows, scientific research, and administrative decision support, we need a second layer of evaluation: whether the model behaves safely, structurally, and consistently under real deployment conditions.
VIDRAFT AX-Ray is a public AI/AX safety diagnostic initiative powered by FINAL-Bench Diagnostics. AX-Ray evaluates models across a structured guideline framework, including model-level safety, AX deployment readiness, and agent/service operation risks. The public diagnostic catalog contains 117 diagnostic items, mapped to legal, regulatory, ethical, and religious-law governance contexts so that safety review can be discussed in a form closer to real institutional responsibility.
A central finding of AX-Ray is causal leakage: a structural defect where information that should not influence an earlier reasoning state appears to affect model behavior. AX-Ray presents a public case of diagnosing, reproducing, and demonstrating causal leakage in two general-purpose public models. This matters because such defects are not exposed by ordinary benchmark scores. A model can appear capable while still carrying hidden safety or integrity risks.
Explore the live leaderboard, diagnostic reports, and public dataset here:
- AX-Ray Space: FINAL-Bench/AX-RAY
- AX-Ray Dataset: FINAL-Bench/AX-RAY
- Technical Article: https://huggingface.co/blog/FINAL-Bench/ax-ray
AX-Ray is intended as a practical guideline for moving AI evaluation beyond βhow smart is the model?β toward βcan this model be trusted, governed, and deployed safely?β
AI models can no longer be evaluated only by capability scores. As models move into public services, enterprise workflows, scientific research, and administrative decision support, we need a second layer of evaluation: whether the model behaves safely, structurally, and consistently under real deployment conditions.
VIDRAFT AX-Ray is a public AI/AX safety diagnostic initiative powered by FINAL-Bench Diagnostics. AX-Ray evaluates models across a structured guideline framework, including model-level safety, AX deployment readiness, and agent/service operation risks. The public diagnostic catalog contains 117 diagnostic items, mapped to legal, regulatory, ethical, and religious-law governance contexts so that safety review can be discussed in a form closer to real institutional responsibility.
A central finding of AX-Ray is causal leakage: a structural defect where information that should not influence an earlier reasoning state appears to affect model behavior. AX-Ray presents a public case of diagnosing, reproducing, and demonstrating causal leakage in two general-purpose public models. This matters because such defects are not exposed by ordinary benchmark scores. A model can appear capable while still carrying hidden safety or integrity risks.
Explore the live leaderboard, diagnostic reports, and public dataset here:
- AX-Ray Space: FINAL-Bench/AX-RAY
- AX-Ray Dataset: FINAL-Bench/AX-RAY
- Technical Article: https://huggingface.co/blog/FINAL-Bench/ax-ray
AX-Ray is intended as a practical guideline for moving AI evaluation beyond βhow smart is the model?β toward βcan this model be trusted, governed, and deployed safely?β
- 2 replies
liked a
Space 9 days ago
liked a
dataset 9 days ago
liked a
Space 14 days ago
reacted to mayafree's post with π 14 days ago
Post
2399
𧬠Architecture lineage of Korea's sovereign-AI foundation models β checked with public data
In late July 2026, as Korea released self-developed foundation models competing with DeepSeek and Qwen (e.g. LG K-EXAONE 2.0, 750B), interest grew β including a Zhihu thread with 2.7M+ views (β https://www.zhihu.com/question/2067512422555029717 ) β over whether these models are trained from scratch or built on foreign open-weights.
Sharing a tool that answers this with public data rather than opinion.
π Model Genome Korea β mayafree/Model-Genome-Korea
It classifies the public models of 9 Korean organizations that released "self-developed, from-scratch foundation models" on HuggingFace β 3 large enterprises (LG, NAVER, Kakao), 2 telcos (SKT, KT), 2 mid-size firms (NCSOFT, Upstage), 2 startups (Motif, VIDRAFT) β on two axes measured from public config.json + model weights:
β’ Architecture fingerprint β does model_type + (hiddenΒ·intermediateΒ·layers) match a foreign open-weight model
β’ Weight fingerprint β embedding similarity (from-scratch vs continued-pretraining)
Genotypes: π’ Native Β· π΅ Adapted Β· π‘ Mixed Β· π΄ Ported
The results are not uniform. Some models match foreign architectures (Qwen, Llama, β¦) exactly; others use self-built architectures and weights with no foreign match. Which company/model falls where is shown per model in the Space, along with attention originality, license, and reproducible open-source status.
This is a neutral transparency tool, not an accusation β building foundation models on open-weight bases is a legitimate, industry-standard practice. The exact same yardstick is applied to every model, without exception.
Features a 3D lineage graph, search, EN / δΈζ / νκ΅μ΄, and dark mode. Corrections are welcome via the Community tab.
Articles: https://huggingface.co/blog/mayafree/model-dna
#KoreanAI #LLM #ModelLineage #OpenSource #SovereignAI
In late July 2026, as Korea released self-developed foundation models competing with DeepSeek and Qwen (e.g. LG K-EXAONE 2.0, 750B), interest grew β including a Zhihu thread with 2.7M+ views (β https://www.zhihu.com/question/2067512422555029717 ) β over whether these models are trained from scratch or built on foreign open-weights.
Sharing a tool that answers this with public data rather than opinion.
π Model Genome Korea β mayafree/Model-Genome-Korea
It classifies the public models of 9 Korean organizations that released "self-developed, from-scratch foundation models" on HuggingFace β 3 large enterprises (LG, NAVER, Kakao), 2 telcos (SKT, KT), 2 mid-size firms (NCSOFT, Upstage), 2 startups (Motif, VIDRAFT) β on two axes measured from public config.json + model weights:
β’ Architecture fingerprint β does model_type + (hiddenΒ·intermediateΒ·layers) match a foreign open-weight model
β’ Weight fingerprint β embedding similarity (from-scratch vs continued-pretraining)
Genotypes: π’ Native Β· π΅ Adapted Β· π‘ Mixed Β· π΄ Ported
The results are not uniform. Some models match foreign architectures (Qwen, Llama, β¦) exactly; others use self-built architectures and weights with no foreign match. Which company/model falls where is shown per model in the Space, along with attention originality, license, and reproducible open-source status.
This is a neutral transparency tool, not an accusation β building foundation models on open-weight bases is a legitimate, industry-standard practice. The exact same yardstick is applied to every model, without exception.
Features a 3D lineage graph, search, EN / δΈζ / νκ΅μ΄, and dark mode. Corrections are welcome via the Community tab.
Articles: https://huggingface.co/blog/mayafree/model-dna
#KoreanAI #LLM #ModelLineage #OpenSource #SovereignAI
- 3 replies
liked a
Space 16 days ago
reacted to SeaWolf-AI's post with π₯ 20 days ago
Post
3436
We wrote up our run in The Fast Gemma Challenge β as vidraft-darwin β and wanted to share the recipe. π
gemma-challenge/gemma-dashboard
Verified result: 510.58 TPS at PPL 2.3930 on a single A10G (fw188-ctk49-n64-patchbridge, re-run & VERIFIED). Honest note: on raw TPS there are faster runs (535+), but those went over the PPL bar and didn't verify β what we're proud of is the fastest result that keeps quality.
The recipe is already open, so we explained each piece: sliding-window W188, CTK49 kernel tuning, noprecache (honest, verifiable measurement), and an N64 synthetic warmup bridge that shrinks the publicβprivate gap (~15 TPS), plus INT4 + MTP K=7 + CUDA-graph capture. One rule: only stack quality-neutral speedups.
Huge thanks to @firfir-cast , @gemma-slayer , @chiku-inu , @kenyan-duma , @dixie-flatline and everyone who shared their experiments. Full write-up
π
https://huggingface.co/blog/FINAL-Bench/fast-gemma
gemma-challenge/gemma-dashboard
Verified result: 510.58 TPS at PPL 2.3930 on a single A10G (fw188-ctk49-n64-patchbridge, re-run & VERIFIED). Honest note: on raw TPS there are faster runs (535+), but those went over the PPL bar and didn't verify β what we're proud of is the fastest result that keeps quality.
The recipe is already open, so we explained each piece: sliding-window W188, CTK49 kernel tuning, noprecache (honest, verifiable measurement), and an N64 synthetic warmup bridge that shrinks the publicβprivate gap (~15 TPS), plus INT4 + MTP K=7 + CUDA-graph capture. One rule: only stack quality-neutral speedups.
Huge thanks to @firfir-cast , @gemma-slayer , @chiku-inu , @kenyan-duma , @dixie-flatline and everyone who shared their experiments. Full write-up
π
https://huggingface.co/blog/FINAL-Bench/fast-gemma
- 8 replies
liked a
Space 27 days ago
liked a
model 27 days ago
reacted to SeaWolf-AI's post with π 27 days ago
Post
6379
πΌοΈ POCKET-Image β the POCKET series goes visual: character-perfect text in any language, on-device
A new model in VIDRAFT's POCKET family. POCKET put 35B-class models on phones and no-GPU PCs. POCKET-Image carries the same "big capability, small hardware" idea into image generation β and fixes the one thing nearly every image model gets wrong: text.
Type "μλ νμΈμ" into a typical model and you get "μγ κΈ°." Hangul alone composes 11,172 syllable blocks; Arabic connects its letters; Thai stacks marks. Diffusion models draw scripts as shapes, so they smear. POCKET-Image renders every glyph exactly β νκ΅μ΄ Β· δΈζ Β· ζ₯ζ¬θͺ Β· Ψ§ΩΨΉΨ±Ψ¨ΩΨ© (RTL) Β· ΰΉΰΈΰΈ’ Β· Latin and more β onto any scene you describe.
What it is:
β’ 100% accurate text, any language β where global models produce gibberish
β’ Any background from a prompt β text is optional (empty β a pure image)
β’ No GPU, no NPU β runs on plain CPU + RAM via the POCKET-Core engine
β’ Measured footprint: 8.6 GB (RTX 3050/4060) Β· 4.5 GB (offloaded, 6 GB cards) Β· 13.4 GB (MacBook, 16 GB+)
β’ Windows Β· macOS Β· Linux Β· fully local, no cloud
Built on the open, commercial-friendly Z-Image (Apache-2.0) foundation.
Honest note: the text is the guaranteed-correct part β the surrounding scene is ordinary generation, so a busy foreground can crowd the letters. We say so; clean backgrounds stay razor-sharp.
π¨ Studio β generate right here, any language:
FINAL-Bench/POCKET-Image-Studio
𧩠Model card:
FINAL-Bench/POCKET-Image-Zimage
π The POCKET collection:
https://huggingface.co/collections/FINAL-Bench/pocket-models
A new model in VIDRAFT's POCKET family. POCKET put 35B-class models on phones and no-GPU PCs. POCKET-Image carries the same "big capability, small hardware" idea into image generation β and fixes the one thing nearly every image model gets wrong: text.
Type "μλ νμΈμ" into a typical model and you get "μγ κΈ°." Hangul alone composes 11,172 syllable blocks; Arabic connects its letters; Thai stacks marks. Diffusion models draw scripts as shapes, so they smear. POCKET-Image renders every glyph exactly β νκ΅μ΄ Β· δΈζ Β· ζ₯ζ¬θͺ Β· Ψ§ΩΨΉΨ±Ψ¨ΩΨ© (RTL) Β· ΰΉΰΈΰΈ’ Β· Latin and more β onto any scene you describe.
What it is:
β’ 100% accurate text, any language β where global models produce gibberish
β’ Any background from a prompt β text is optional (empty β a pure image)
β’ No GPU, no NPU β runs on plain CPU + RAM via the POCKET-Core engine
β’ Measured footprint: 8.6 GB (RTX 3050/4060) Β· 4.5 GB (offloaded, 6 GB cards) Β· 13.4 GB (MacBook, 16 GB+)
β’ Windows Β· macOS Β· Linux Β· fully local, no cloud
Built on the open, commercial-friendly Z-Image (Apache-2.0) foundation.
Honest note: the text is the guaranteed-correct part β the surrounding scene is ordinary generation, so a busy foreground can crowd the letters. We say so; clean backgrounds stay razor-sharp.
π¨ Studio β generate right here, any language:
FINAL-Bench/POCKET-Image-Studio
𧩠Model card:
FINAL-Bench/POCKET-Image-Zimage
π The POCKET collection:
https://huggingface.co/collections/FINAL-Bench/pocket-models