The image model of Geode's Images screen: a converted copy of FLUX.2 [klein] 4B by Black Forest Labs (a 4-step distilled rectified-flow transformer, 3.9 B parameters, with the Qwen3-4B text encoder), split into one file per stage so the phone loads the text encoder, runs it and frees it before the transformer loads, and the transformer before the VAE (only the largest stage is ever in memory). Geode's own engine runs it on the phone's processor, graphics chip or Snapdragon NPU; nothing leaves the phone.
Two folders; the app downloads one:
| Folder | Transformer (DiT) | Download | Quality against the original (fp32, seed 42, 512²) |
|---|---|---|---|
fp16/ (default) |
fp16, 7.2 GiB | 10.2 GB | the same image (76 dB PSNR, SSIM 1.000) |
int8/ (opt-in, lower quality) |
int8 per 32 inputs with a Hadamard-32 turn ("i8r"), 4.0 GiB | 6.7 GB | the same picture with fine texture differing (30-41 dB, SSIM 0.97-0.997) |
Both share the same text encoder (Q4), VAE (fp16) and tokenizer files (byte for byte).
| File | What it is |
|---|---|
tokenizer.bin |
Qwen3's byte-level BPE (the checkpoint's tokenizer/) in Geode's format |
embed.bin |
the text encoder's token embeddings, bf16 as in the checkpoint (one row read per token) |
te_q4.bin |
the text encoder's layers 0-26 in 4 bits (an E4M3 scale per 16 values); its norms fp32 |
dit_f16.bin / dit_i8r.bin |
the transformer: fp16, or its blocks' matrices in int8 (the small modulation / embedding / output layers stay fp16) |
vae_f16.bin |
the VAE's decoder and the latents' batch-norm statistics, fp16 (norms and biases fp32) |
vae_enc_f16.bin |
the VAE's encoder (+ quant_conv), fp16: reference images for editing (added 2026-11-18) |
taef2_f16.bin |
TAEF2's tiny decoder (madebyollin/taef2, MIT), fp16: the live preview after each step (added 2026-11-18) |
vae_small_f16.bin |
BFL's FLUX.2 small decoder (Apache-2.0), fp16: an opt-in faster decoder, downloaded only when turned on; its own pack in geode.json's vae_small (added 2026-11-18) |
geode.json |
the pack's manifest: the configuration Geode reads, every file's size and SHA-256, every tensor |
LICENSE |
the Apache License 2.0, as the original model ships it |
LICENSE-TAEF2 |
TAEF2's MIT licence (Copyright (c) 2023 Ollin Boer Bohan) |
e7b7dc27f91deacad38e78976d1f2b499d76a294, Copyright Black Forest Labs, licensed under the Apache License 2.0
(LICENSE copied here unchanged).tools/pack_image.py):fp16/; relative error ~1e-8), or its blocks' matrices and
context_embedder quantized to int8 with one fp16 scale per 32 inputs after a normalized 32-point Walsh-Hadamard
transform of each group (int8/);vae_enc_f16.bin;vae_small_f16.bin, unchanged otherwise;LICENSE-TAEF2): its decoder as fp16
rows in taef2_f16.bin;Flux2KleinPipeline, so a seed gives the original pipeline's image up to the
precision above.Black Forest Labs' model card lists the uses the model and its derivatives are not meant for: among them anything that violates applicable law, exploiting or harming minors, deceptive or fraudulent content, harassment, and non-consensual intimate imagery. Those apply to these files too. Nothing here changes the licence or adds terms of its own.
Measured per stage by Geode's engine and shown in the app (Images screen, the result's details); the targets are 512² in about 10 s on a Snapdragon 8 Elite (the NPU) and 2-3 minutes on a Snapdragon 720G (the processor).
fp16/dit_f16.bin 7.8 GBint8/dit_i8r.bin 4.3 GBfp16/te_q4.bin 1.5 GBint8/te_q4.bin 1.5 GBfp16/embed.bin 778 MBint8/embed.bin 778 MBfp16/vae_f16.bin 99 MBint8/vae_f16.bin 99 MBfp16/vae_enc_f16.bin 69 MBint8/vae_enc_f16.bin 69 MBfp16/vae_small_f16.bin 56 MBint8/vae_small_f16.bin 56 MBfp16/tokenizer.bin 3 MBint8/tokenizer.bin 3 MBfp16/taef2_f16.bin 3 MBint8/taef2_f16.bin 3 MBblack-forest-labs/FLUX.2-klein-4B