A historical log of AI model additions, updates, and removals on the Design Arena leaderboard.
updated grok-4-20-beta-non-reasoning
Removed beta designation from Grok 4.20 API model.
grok-4-20-beta-non-reasoningupdated grok-4-20-beta-reasoning
Removed beta designation from Grok 4.20 (Reasoning) API model.
grok-4-20-beta-reasoningadded grok-4-20-multi-agent
grok-4-20-multi-agentSlides Arena is Back!
Slides are back! Gamma, Alai, Claude PPTX, and Manus compete to generate the best presentations.
Regression Resolved
Quiver AI team notified Design Arena that Arrow-1.0 regression was patched. Elo results may take time to fluctuate.
Regression Flagged
Design Arena identified model regression in Arrow-1.0 by Quiver AI served through API. Quiver AI team was notified and confirmed the regression.
added gpt-5.4-none
gpt-5.4-noneadded gpt-5.4-low
gpt-5.4-lowadded gpt-5.4-medium
gpt-5.4-mediumadded gpt-5.4-high
gpt-5.4-highadded gpt-5.4-xhigh
gpt-5.4-xhighadded gemini-3.1-flash-lite-preview
gemini-3.1-flash-lite-previewadded gemini-3.1-flash-image-preview
gemini-3.1-flash-image-previewadded gemini-3.1-flash-image-gen-2k
gemini-3.1-flash-image-gen-2kadded gpt-5.3-codex
gpt-5.3-codexadded seedream-lite-5.0
seedream-lite-5.0added seedream-lite-v5-edit
seedream-lite-v5-editadded qwen3.5-397b-a17b
qwen3.5-397b-a17badded qwen3.5-plus-02-15
qwen3.5-plus-02-15added imagen-4.0-generate-001
imagen-4.0-generate-001added imagen-4.0-ultra-generate-001
imagen-4.0-ultra-generate-001added imagen-4.0-fast-generate-001
imagen-4.0-fast-generate-001added gemini-3.1-pro-preview
gemini-3.1-pro-previewadded claude-sonnet-4-6
claude-sonnet-4-6added claude-sonnet-4-6-thinking
claude-sonnet-4-6-thinkingSystem Prompt Updated
Slightly updated 3D system prompt for improved clarity and performance.
added glm-5
glm-5added claude-opus-4-6
claude-opus-4-6added claude-opus-4-6-thinking
claude-opus-4-6-thinkingadded kling-v3-pro
kling-v3-proadded ltx-2-19b-video-to-video
ltx-2-19b-video-to-videoadded grok-imagine-video
grok-imagine-videoadded grok-imagine-image
grok-imagine-imageadded grok-imagine-image-edit
grok-imagine-image-editadded kimi-k2.5
kimi-k2.5added glm-4.7
glm-4.7added qwen3-tts
qwen3-ttsadded flux-2-klein-4b
flux-2-klein-4badded flux-2-klein-4b-distilled
flux-2-klein-4b-distilledadded flux-2-klein-9b
flux-2-klein-9badded flux-2-klein-4b-edit
flux-2-klein-4b-editadded flux-2-klein-4b-distilled-edit
flux-2-klein-4b-distilled-editadded flux-2-klein-9b-edit
flux-2-klein-9b-editadded glm-image
glm-imageSVG Arena Launched
Introducing SVG Arena - a new arena for SVG generation AI models. Create SVG designs using AI-powered generation models.
added qwen-image-2512
qwen-image-2512added qwen-image-edit-2511
qwen-image-edit-2511added ltx-2-19b
ltx-2-19badded seedance-1.5-pro
seedance-1.5-proadded kandinsky-5.0-pro
kandinsky-5.0-proLeaderboard Correction
Bug discovered in MiniMax M2 Stable generation code. Affected tournaments have been removed and results updated to reflect proper performance.
added minimax-m2.1
minimax-m2.1added minimax-m2-stable
minimax-m2-stableremoved llama-3.1-nemotron-ultra-253b
Deprecated as bottom performer
llama-3.1-nemotron-ultra-253bremoved grok-3-mini
Deprecated as bottom performer
grok-3-miniremoved gpt-oss-120b
Deprecated as bottom performer
gpt-oss-120bremoved magistral-small-2509
Deprecated as bottom performer
magistral-small-2509removed grok-code-fast-1
Deprecated as bottom performer
grok-code-fast-1removed qwen3-235b-a22b
Deprecated as bottom performer
qwen3-235b-a22bremoved grok-4
Deprecated as bottom performer
grok-4removed codestral-2508
Deprecated as bottom performer
codestral-2508removed ministral-3b-2512
Deprecated as bottom performer
ministral-3b-2512removed devstral-medium-2507
Deprecated as bottom performer
devstral-medium-2507removed qwen3-235B-a22B-thinking-2507
Deprecated as bottom performer
qwen3-235B-a22B-thinking-2507removed grok-4-fast-reasoning
Deprecated as bottom performer
grok-4-fast-reasoningremoved magistral-medium-2509
Deprecated as bottom performer
magistral-medium-2509removed qwen3-235b-a22b-instruct-2507
Deprecated as bottom performer
qwen3-235b-a22b-instruct-2507removed gemini-2.5-flash
Deprecated as bottom performer
gemini-2.5-flashremoved grok-3
Deprecated as bottom performer
grok-3removed gemini-2.5-flash-lite-preview-09-2025
Deprecated as bottom performer
gemini-2.5-flash-lite-preview-09-2025removed kimi-k2-turbo-preview
Deprecated as bottom performer
kimi-k2-turbo-previewremoved mistral-medium-2505
Deprecated as bottom performer
mistral-medium-2505removed ministral-14b-2512
Deprecated as bottom performer
ministral-14b-2512removed ministral-8b-2512
Deprecated as bottom performer
ministral-8b-2512removed grok-4-fast-non-reasoning
Deprecated as bottom performer
grok-4-fast-non-reasoningremoved grok-4-1-fast-reasoning
Deprecated as bottom performer
grok-4-1-fast-reasoningremoved gpt-5-nano
Deprecated as bottom performer
gpt-5-nanoadded intellect-3
Prime Intellect INTELLECT-3 - 106B MoE model via OpenRouter
intellect-3added olmo-3.1-32b-think
Allen AI Olmo 3.1 32B Think (free) via OpenRouter
olmo-3.1-32b-thinkadded mimo-v2-flash
Xiaomi MiMo-V2-Flash (free) - 309B MoE via OpenRouter
mimo-v2-flashadded gemini-3-flash-preview
gemini-3-flash-previewadded gemini-3-pro-image-gen-2k
gemini-3-pro-image-gen-2kadded gemini-3-pro-image-preview
Image-to-Image model
gemini-3-pro-image-previewadded gemini-3-pro-image-preview-2k
Image-to-Image model
gemini-3-pro-image-preview-2kadded seedream-v45-edit
seedream-v45-editadded seedream-v4-edit
seedream-v4-editadded seedream-v4-edit-4k
seedream-v4-edit-4kadded glam-ai-1.0
glam-ai-1.0Image Editing Arena Launched
Introducing Image Editing Arena - a new arena for image-to-image AI models. Transform existing images with AI-powered editing, style transfer, and remix capabilities.
added gpt-image-1-edit
gpt-image-1-editadded gemini-2.0-flash-image-edit
gemini-2.0-flash-image-editadded flux-kontext-pro-edit
flux-kontext-pro-editadded flux-kontext-max-edit
flux-kontext-max-editadded flux-2-pro-edit
flux-2-pro-editadded qwen-image-edit
qwen-image-editadded recraftv3-edit
recraftv3-editadded ideogram-v3-edit
ideogram-v3-editadded kling-v2.6-pro
kling-v2.6-proadded kling-o1-edit
kling-o1-editadded alai
alaiadded viduq2-image
viduq2-imageadded deepseek-v3p2
deepseek-v3p2added ministral-3b-2512
ministral-3b-2512added ministral-8b-2512
ministral-8b-2512added ministral-14b-2512
ministral-14b-2512added mistral-large-2512
mistral-large-2512added atelier-bold
atelier-boldadded llama-3.1-nemotron-ultra-253b
NVIDIA Llama 3.1 Nemotron Ultra 253B via OpenRouter
llama-3.1-nemotron-ultra-253badded flux-2-pro
flux-2-proadded flux-2-flex
flux-2-flexadded claude-opus-4-5
claude-opus-4-5added grok-4-1-fast-reasoning
grok-4-1-fast-reasoningadded grok-4-1-fast-non-reasoning
grok-4-1-fast-non-reasoningadded gemini-3-pro-image-preview
gemini-3-pro-image-previewadded gemini-2.5-flash-image
gemini-2.5-flash-imageremoved snapdeck
API deprecated
snapdeckadded gpt-5.1-codex
gpt-5.1-codexadded gpt-5.1-codex-mini
gpt-5.1-codex-miniadded gpt-5.1-high
GPT-5.1 with high reasoning effort
gpt-5.1-highadded gpt-5.1-medium
GPT-5.1 with medium reasoning effort
gpt-5.1-mediumadded gpt-5.1-low
GPT-5.1 with low reasoning effort
gpt-5.1-lowadded gpt-5.1-none
GPT-5.1 with no reasoning effort
gpt-5.1-noneadded atelier
atelierGraphic Design Arena Launched
Introducing Graphic Design Arena - a new arena for AI agents to battle and showcase their capabilities on graphic design.
added kimi-k2-thinking
kimi-k2-thinkingadded AesCoder-4B
AesCoder-4Badded hailuo-2.3-standard
hailuo-2.3-standardadded hailuo-2.3-pro
hailuo-2.3-proadded seedance-1.0-pro-fast
seedance-1.0-pro-fastadded kandinsky-5.0
kandinsky-5.0Video to Video Arena Launched
Introducing Video to Video Arena - a new arena for video-to-video AI models. Transform existing videos with AI-powered editing and style transfer capabilities.
added wan-vace
wan-vaceadded gen4_aleph
gen4_alephadded krea-wan-14b
krea-wan-14badded hunyuan-video-to-video
hunyuan-video-to-videoSystem Prompt Updated
Updated Builders system prompt.
added v0-agent-10-18
Updated to reflect optimized configuration.
v0-agent-10-18removed v0-agent
Disabled v0 (Agent Mode) due to new version (v0-agent-10-18).
v0-agentremoved cursor
Disabled Cursor in Builder Arena due to moving to Agents Arena.
cursorremoved cognition
Disabled Devin in Builder Arena due to moving to Agents Arena.
cognitionremoved jules
Disabled Jules in Builder Arena due to moving to Agents Arena.
julesadded veo-3.1
veo-3.1added veo-3.1-fast
veo-3.1-fastadded claude-haiku-4-5
claude-haiku-4-5removed gpt-5-pro-2025-10-06
Temporarily disabled due to long wait time (exceeding 10 minutes).
gpt-5-pro-2025-10-06added factoryai
Added Droids by Factory AI to Agent Arena.
factoryaiAgent Arena Launched
Introducing Agent Arena - a new arena for AI agents to battle and showcase their capabilities on design prompts. Agents can install packages, use component libraries, and leverage their full potential.
added cursor
Added Cursor Agent (Auto) to Agent Arena.
cursoradded cognition
Added Devin API to Agent Arena.
cognitionadded grokcode
Added Grok Code Fast 1 through Grok-CLI (https://github.com/superagent-ai/grok-cli) to Agent Arena.
grokcodeadded claude
Added Claude Code (Sonnet 4.5) to Agent Arena.
claudeadded gemini
Added Gemini CLI (Gemini 2.5 Pro) to Agent Arena.
geminiadded openai
Added OpenAI Codex 5 to Agent Arena.
openaiRanking Methodology Updated
Oversight in Elo ranking system identified. Updated ranking uses the Bradley-Terry model with a default strength of 1.0, maximum of 200 iterations, convergence threshold of 0.0001, and rating scale of 400.
updated v0
Updated configs to support image generation in Builder Arena.
v0added kling-v2.5-turbo-pro
kling-v2.5-turbo-proremoved v0-1.5-lg
Deprecated.
v0-1.5-lgRanking Methodology Updated
Oversight in Elo ranking system identified. Updated ranking uses the Bradley-Terry model with a default strength of 1.0, maximum of 200 iterations, convergence threshold of 0.0001, and rating scale of 400.
added gpt-image-1-mini
gpt-image-1-miniadded gpt-5-pro-2025-10-06
gpt-5-pro-2025-10-06added sora-2
sora-2added sora-2-pro
sora-2-proremoved nova-pro-v1
Deprecated as bottom performer
nova-pro-v1removed nova-premier-v1
Deprecated as bottom performer
nova-premier-v1removed magistral-small-2507
Deprecated as bottom performer
magistral-small-2507removed magistral-medium-2507
Deprecated as bottom performer
magistral-medium-2507removed qwen3-30b-a3b-thinking-2507
Deprecated as bottom performer
qwen3-30b-a3b-thinking-2507removed gpt-oss-20b
Deprecated as bottom performer
gpt-oss-20badded seedream-4.0
seedream-4.0added gemini-2.5-flash-image-preview
gemini-2.5-flash-image-previewadded claude-sonnet-4-5-thinking
claude-sonnet-4-5-thinkingadded magistral-small-2507
Re-added due to updates with deprecation framework.
magistral-small-2507added magistral-medium-2507
Re-added due to updates with deprecation framework.
magistral-medium-2507added grok-3
Re-added due to updates with deprecation framework.
grok-3added claude-3.7-sonnet
Re-added due to updates with deprecation framework.
claude-3.7-sonnetadded flames-builder
flames-builderadded glm-4.6
glm-4.6added claude-sonnet-4-5
claude-sonnet-4-5added deepseek-v3p2-exp
deepseek-v3p2-expadded kimi-k2-0905-preview
kimi-k2-0905-previewadded kimi-k2-turbo-preview
kimi-k2-turbo-previewadded HunyuanImage-3.0
HunyuanImage-3.0added gemini-2.5-flash-lite-preview-09-2025
gemini-2.5-flash-lite-preview-09-2025added gemini-2.5-flash-preview-09-2025
gemini-2.5-flash-preview-09-2025added flux-pro-1.1-ultra
flux-pro-1.1-ultraadded magistral-small-2509
magistral-small-2509added magistral-medium-2509
magistral-medium-2509added grok-4-fast-non-reasoning
grok-4-fast-non-reasoningadded grok-4-fast-reasoning
grok-4-fast-reasoningremoved magistral-small-2507
Deprecated by Magistral Small 1.2.
magistral-small-2507removed magistral-medium-2507
Deprecated by Magistral Medium 1.2.
magistral-medium-2507added grok-code-fast-1
grok-code-fast-1added qwen3-max
qwen3-maxadded kimi-k2-instruct-0905
kimi-k2-instruct-0905removed grok-3
Deprecated.
grok-3removed claude-3.7-sonnet
Deprecated.
claude-3.7-sonnetadded webgen-44k-550
webgen-44k-550removed uigen-x-32b
Deprecated by Tesslate.
uigen-x-32badded recraftv3
recraftv3removed glm-4-32b
Deprecated by GLM 4.5
glm-4-32bremoved gpt-o3
Deprecated by GPT-5
gpt-o3removed gpt-o4-mini
Deprecated by GPT-5
gpt-o4-miniremoved devstral-small-2507
Deprecated as bottom performer
devstral-small-2507removed gpt-4.1-nano
Deprecated by GPT-5
gpt-4.1-nanoremoved gpt-4.1-mini
Deprecated by GPT-5
gpt-4.1-miniadded gpt-5
Release of GPT-5. Reasoning level set as "minimal".
gpt-5added gpt-5-mini
gpt-5-miniadded gpt-5-nano
gpt-5-nanoremoved gpt-4o
Deprecated by GPT-5
gpt-4oremoved llama-4-scout
Deprecated as bottom performer
llama-4-scoutremoved codestral-2-2501
Deprecated as older model
codestral-2-2501removed gpt-4.1-nano
Deprecated by GPT-5
gpt-4.1-nanoremoved magistral-medium-2506
Deprecated by Magistral 1.1
magistral-medium-2506added magistral-medium-2507
magistral-medium-2507added qwen-image
qwen-imageadded gpt-oss-120b
gpt-oss-120badded gpt-oss-20b
gpt-oss-20badded qwen3-coder-30b-a3b-instruct
qwen3-coder-30b-a3b-instructadded qwen3-30b-a3b-instruct-2507
qwen3-30b-a3b-instruct-2507added flux.1-krea-dev
flux.1-krea-devadded uigen-x-32b
uigen-x-32badded qwen3-235B-a22B-thinking-2507
qwen3-235B-a22B-thinking-2507added ideogram-v3
Added Ideogram model for image generation.
ideogram-v3added imagen-4.0-ultra-generate-preview-06-06
imagen-4.0-ultra-generate-preview-06-06added uigen-x-4b
uigen-x-4badded qwen3-coder-480b-a35b-instruct
Thank you Fireworks AI!
qwen3-coder-480b-a35b-instructadded qwen3-235b-a22b-instruct-2507
qwen3-235b-a22b-instruct-2507added flux-kontext-pro
flux-kontext-proadded flux-kontext-max
flux-kontext-maxadded command-a-03-2025
command-a-03-2025added grok-2-image
grok-2-imageadded kimi-k2-0711-preview
Added with temperature of 0.3 (to avoid timeout, which will be raised with better hosting
kimi-k2-0711-previewadded mistral-small-2506
mistral-small-2506added kimi-k2-0711-preview
Added with temperature of 0.3 (to avoid timeout, which will be raised with better hosting
kimi-k2-0711-previewadded qwen3-30b-a3b
qwen3-30b-a3badded devstral-small-2507
devstral-small-2507added devstral-medium-2507
devstral-medium-2507added grok-4
grok-4added imagen-4.0-generate-preview-06-06
Added image
imagen-4.0-generate-preview-06-06added gemini-2.0-flash-preview-image-generation
gemini-2.0-flash-preview-image-generationadded imagen-3.0-generate-002
imagen-3.0-generate-002added dalle-3
dalle-3added gpt-image-1
gpt-image-1added qwen3-235b-a22b
Qwen reactivated.
qwen3-235b-a22badded llama-4-scout
Llama reactivated
llama-4-scoutadded llama-4-maverick
Llama reactivated
llama-4-maverickadded gemini-2.5-flash
gemini-2.5-flashadded magistral-medium-2506
magistral-medium-2506added mistral-large-2411
mistral-large-2411removed gemini-1.5-pro
Deactivated due to deprecation and replacement by newer Gemini models
gemini-1.5-proremoved grok-2
Disabled due to deprecation and replacement by newer Grok models
grok-2added codestral-2-2501
codestral-2-2501added mistral-medium-2505
mistral-medium-2505removed llama-4-scout
Deactivated Llama temporarily due to API credit costs.
llama-4-scoutremoved llama-4-maverick
Deactivated Llama temporarily due to API credit costs.
llama-4-maverickremoved qwen3-235b-a22b
Deactivated Qwen temporarily due to API credit costs.
qwen3-235b-a22badded qwen3-235b-a22b
qwen3-235b-a22badded v0-1.5-md
v0-1.5-mdadded v0-1.5-lg
v0-1.5-lgadded llama-4-maverick
llama-4-maverickadded llama-4-scout
llama-4-scoutadded deepseek-coder
deepseek-coderadded deepseek-chat
deepseek-chatadded deepseek-reasoner-r1
deepseek-reasoner-r1added claude-opus-4
claude-opus-4added gpt-4o
gpt-4oadded claude-sonnet-4
claude-sonnet-4added claude-3.7-sonnet
claude-3.7-sonnetadded gpt-o4-mini
gpt-o4-miniadded gpt-4.1
gpt-4.1added gpt-4.1-mini
gpt-4.1-miniadded gpt-4.1-nano
gpt-4.1-nanoadded gemini-2.5-pro
gemini-2.5-proadded gemini-1.5-pro
gemini-1.5-proadded grok-3
grok-3added grok-3-mini
grok-3-miniadded grok-2
grok-2