The Open Weights
LatestModelsLeaderboardsCompanies
Subscribe
The Open Weights

The daily record of open-source AI. New model releases, leaderboards, and what's coming next — written for people who ship.

Refreshed every 12 hours

Discover

  • Latest releases
  • New today
  • Trending models

Browse

  • All models
  • Companies
  • Categories
  • Leaderboards

About

  • About
  • Editorial policy
  • Hardware estimates
  • RSS feed
  • llms.txt
  • Newsletter

© 2026 The Open Weights. An independent publication.

PrivacyTermsSMSAggregated by Claude · curated by humans.
LatestOpenMOSSBase
OpenMOSSText → Image

OpenMOSS Releases Apache-2.0 Bilingual Image Model

MOSS-Image Base arrives as a permissively licensed text-to-image foundation with English and Chinese support.

Oct 11, 2026
UpdateApache 2.0
MOSS-Image SFT

OpenMOSS has released MOSS-Image Base, a text-to-image model published under the permissive Apache-2.0 license. Positioned as a base model, it is aimed at developers and researchers who want a foundation to build on rather than a finished consumer product.

The standout detail is its bilingual design: the model is built to handle prompts in both English and Chinese. That broadens its usefulness for teams working across both languages, a practical gap that many Western-origin image models leave open.

Why it matters

Apache-2.0 licensing is the headline here. Unlike some image generators that ship with usage restrictions or research-only terms, a true permissive license gives teams clear latitude to fine-tune, deploy, and build commercial tools on top of the weights.

  • Open Apache-2.0 license with few restrictions
  • Bilingual English and Chinese prompt support
  • Released as a base model for downstream fine-tuning

OpenMOSS has not published parameter counts, supported resolutions, or benchmark results alongside the initial drop, so direct comparisons to established open models remain hard to make. For now, the release is best read as an invitation to experiment, with the community likely to surface its real strengths and limits as people put it through its paces. The weights are available now on Hugging Face.

Sources

  • OpenMOSS-Team/MOSS-Image-Base

    Hugging Face

    Visit
OlderIndra OCR brings CPU-friendly document parsingDatalab To · Vision-Language · last monthNewerOpenMOSS Releases MOSS-Image Under Apache 2.0OpenMOSS · Text → Image · last month

Get the model

Hugging Face

Specs

Size10.2 GB
PrecisionBF16
ArchitectureMOSSImagePipeline
LicenseAPACHE-2.0
Downloads14K
Likes672

Can you run it?

Runs on a laptop — about 6.1 GB at FP8.

  • BF16 (as published)

    12 GB GPU (RTX 3060 / 4070) · 16 GB Mac

    11.2 GB
  • FP8

    8 GB GPU (RTX 4060) · 16 GB laptop or Mac

    6.1 GB
Your machine
  • Covers the published checkpoint; text encoders and VAE loaded alongside may add a few GB (many runtimes offload them to CPU).

Based on Hugging Face weights metadata. Assumes an 8K context and ~1 GB runtime overhead; actual needs vary. How we estimate


Modalities

Text → Image
2 versions — view changelog

The Weekly Weights

Every open release that mattered, one email a week.

0 comments

No comments yet. Be the first to weigh in.

More from OpenMOSS

All OpenMOSS releases →
MOSS-Image SFT
OpenMOSS/Text → ImageRuns on a laptop

OpenMOSS Releases MOSS-Image Under Apache 2.0

The academic team behind the MOSS project moves into text-to-image generation with a permissively licensed model and a fine-tuned SFT variant.

Oct 11, 2026
OpenMOSS/Vision-Language

OpenMOSS Debuts MOSS-VL for Real-Time Vision Interaction

A new open vision-language model family uses gated cross-attention to enable streaming, low-latency multimodal exchanges.

Aug 14, 2026
OpenMOSS/Text / LLMRuns on a laptop

MameLoshnLM Brings Yiddish Into the Open LLM Era

OpenMOSS releases what it calls the first open-source 8B language model built specifically for Yiddish, paired with an evaluation benchmark.

Aug 5, 2026

More in Text → Image

All Text → Image →
MiniMax H3 Image
MiniMax/Text → Image

MiniMax Releases H3 Image, Its Open Text-to-Image Model

The Chinese AI lab enters the open image-generation space with a single model that both creates and edits pictures from text.

Nov 3, 2026
Fibo 2.0-dev
Briaai/Text → ImageRuns on a laptop

Bria's Fibo 2.0 Brings Structured Prompts to Open Image AI

The new text-to-image and editing model leans on structured prompts and multi-reference inputs for more controllable generation.

Oct 18, 2026
FLUX.3 Image Klein
Black Forest Labs/Text → Image

Black Forest Labs ships FLUX.3 Image Klein for ComfyUI

The lab's latest open image model arrives packaged for ComfyUI with both generation and editing built in.

Oct 14, 2026