FLUX 3 models beat FLUX.2 in all 19 Text-to-Image categories; Klein tops $9/1k price frontier
ArtificialAnlys · x · 2026-11-06
Artificial Analysis published a deep evaluation of FLUX 3 Image and FLUX 3 Image Klein on its Image Arena leaderboards.
- Text-to-Image: both models rank above their FLUX.2 predecessors in all 19 categories (9 capabilities + 10 use cases). Layout is a top gain; Klein jumps from bottom third to top quarter on Knowledge and Layout, while FLUX 3 Image enters the top ten for Complex Compositions, Reasoning and Knowledge. Radar charts show the gap to the capability frontier shrinking dramatically versus FLUX.2.
- Image Editing: FLUX 3 Image's biggest use-case gain is UI/UX Design (bottom half → top five); Klein improves most on Productivity & Knowledge Work. Both gain most on the Composition & Framing editing action, with Reasoning-Based Edit next; Identity Preserving Edit remains a weak spot for both.
- Open weights: Klein (5.6B, released Oct 15 under FLUX Kommunity License, free commercial use under $5M annual revenue) is #2 on AA-Image-T2I v2.1 among open weights, neck and neck with Qwen-Image-2.1, and #3 on editing.
- Price: Klein costs $9 per 1k images — no model ranks higher at or below that price; FLUX 3 Image costs $48/1k vs FLUX.2 [max]'s $70 while ranking 22–35 places higher.
More from Models
- Tinker team benchmarks model finetunability: Inkling 2 Small learns faster for less — ziqiao_ma · 2026-11-06
- Inkling 2 Small posts top vision benchmark scores, endorsed by PyTorch creator — tessybarton · 2026-11-06
- 28,849-trial Harness Leaderboard: bash-only mini-swe-agent beats openclaw 51.4% vs 27.3% — bansalg_ · 2026-11-06
- Qwen 27B + Python REPL harness hits 78.9% on Harvey benchmark, beating larger flagships — bansalg_ · 2026-11-06
- Running quantized DeepSeek V4.1 on a single RTX 3090 at 20 tokens/s — GodComplecs · 2026-11-06
- Unsloth's danielhanchen to release quantized versions of Thinking Machines' new model — danielhanchen · 2026-11-06