AI Paper Atlas

Research map / Benchmarks / T2I-CompBench Color

T2I-CompBench Color leaderboard

11 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on T2I-CompBench Color. Best score: 0.8677 (TexTailor (full), ECCV 2026).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

T2I-CompBench Color · score

11 papers, each with its best reported number for its own method; higher is better.

#score ↑MethodPaperTopic
10.8677TexTailor (full)
Qwen Image
TexTailor: Inference-Time Textual Guidance Tailoring for Multimodal Diffusion Transformers
ECCV 2026 · Table 2
Text-to-image diffusion
20.8567Janus-Pro + OSPO
Janus-Pro-7B
OSPO: Object-Centric Self-Improving Preference Optimization for Text-to-Image Generation
CVPR 2026 · Table 1
Text-to-image diffusion
30.838+ ISAC_LO
SD3.5-M
ISAC: Training-Free Instance-to-Semantic Attention Control for Multi-Instance Generation
ECCV 2026 · Table 1
Text-to-image diffusion
40.8335RPG
SDXL
Mastering Text-to-Image Diffusion: Recaptioning, Planning, and Generating with Multimodal LLMs
ICML 2024 · T2I-CompBench table
Text-to-image diffusion
50.806FLUX + IMG
FLUX.1 [dev]
IMG: Calibrating Diffusion Models via Implicit Multimodal Guidance
ICCV 2025 · Table 2
Text-to-image diffusion
60.788CaPO+SD3-M
SD3-M
Calibrated Multi-Preference Optimization for Aligning Diffusion Models
CVPR 2025 · Table 3
Text-to-image diffusion
70.7827CoMat-SDXL
SDXL
CoMat: Aligning Text-to-Image Diffusion Model with Image-to-Text Concept Matching
NeurIPS 2024 · T2I-CompBench result table
Text-to-image diffusion
80.7067IMBA loss from scratch
1B DiT (31M data)
Imbalance in Balance: Online Concept Balancing in Generation Models
ICCV 2025 · Table 2
Text-to-image diffusion
90.6893Ranni (Ours)Ranni: Taming Text-to-Image Diffusion for Accurate Instruction Following
CVPR 2024 · Table 1
Text-to-image diffusion
100.6712AnchorSteer
SDXL
Anchoring and Steering Diffusion: Enhancing the Faithfulness of Text-to-Image Generation at Inference Time
ECCV 2026 · Table 3
Text-to-image diffusion
1 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.