BAGEL
Unified 7B multimodal model that both understands and generates images
About
BAGEL is ByteDance Seed's 7B-active-parameter unified multimodal model, released May 2025 under Apache 2.0. One architecture both understands and generates images and text — generation, editing, style transfer, and visual understanding — using a Mixture-of-Transformer-Experts design. The team reports GenEval 0.88 and MMBench 85.0, ahead of Qwen2.5-VL-7B; longer video understanding remains a weaker spot.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Team-reported: GenEval 0.88 overall; MMBench 85.0; GEdit-Bench 7.36 (single-turn).
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Released 2025-05-21 by ByteDance Seed under Apache 2.0 (confirmed on HF card); 7B active params (14B total MoT); unified understanding + generation; paper arXiv:2505.14683; GenEval 0.88 / MMBench 85.0 figures are team-reported and labeled as such. License: Apache-2.0 confirmed on the Hugging Face card..
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 8.0/10
- Maintenance · recency, releases, issue load
- 8.0/10
- Openness · license + docs
- 10.0/10
- Manual review · sanity read + red-flag check
- 7.0/10
- · Released 2025-05-21 by ByteDance Seed under Apache 2.0 (confirmed on HF card)
- · 7B active params (14B total MoT); unified understanding + generation; paper arXiv:2505.14683
- · GenEval 0.88 / MMBench 85.0 figures are team-reported and labeled as such