Text & Chat Models (LLMs)
Google (Gemini & Gemma)
- gemini-2.5-pro
- gemini-2.5-pro-grounding-exp
- gemini-2.5-flash
- gemini-2.5-flash-preview-09-2025
- gemini-2.5-flash-lite-preview-09-2025-no-thinking
- gemini-2.5-flash-lite-preview-06-17-thinking
- gemini-3-pro
- gemini-2.0-flash-001
- gemma-3-27b-it
- gemma-3n-e4b-it
OpenAI (GPT & O-Series)
- gpt-5.1 / gpt-5.1-high
- gpt-5-chat
- gpt-5-high / gpt-5-high-new-system-prompt / gpt-5-high-no-system-prompt
- gpt-5-mini-high / gpt-5-nano-high
- chatgpt-4o-latest-20250326
- gpt-4.1-2025-04-14 / gpt-4.1-mini-2025-04-14
- gpt-oss-120b / gpt-oss-20b
- o3-2025-04-16 / o3-mini
- o4-mini-2025-04-16
Anthropic (Claude)
- claude-3-7-sonnet-20250219 (+ thinking/thinking-32k)
- claude-3-5-sonnet-20241022
- claude-3-5-haiku-20241022
- claude-opus-4-5-20251101 (+ thinking-32k)
- claude-sonnet-4-5-20250929 (+ thinking-32k)
- claude-haiku-4-5-20251001
- claude-opus-4-1-20250805 (+ thinking-16k)
- claude-opus-4-20250514 (+ thinking-16k)
- claude-sonnet-4-20250514 (+ thinking-32k)
xAI (Grok)
- grok-3-mini-beta / grok-3-mini-high
- grok-4.1 / grok-4.1-thinking
- grok-4-1-fast-reasoning / grok-4-1-fast-non-reasoning
- grok-4-0709
- grok-4-fast-chat / grok-4-fast-reasoning
Alibaba (Qwen)
- qwen3-max-2025-09-23 / 09-26 / 10-20
- qwen3-max-preview / qwen3-max-thinking
- qwen3-next-80b-a3b-instruct / thinking
- qwen3-235b-a22b (+ instruct/thinking/no-thinking)
- qwen3-30b-a3b (+ instruct)
- qwen3-coder-480b-a35b-instruct
- qwen3-omni-flash
- qwq-32b
- Vision Understanding: qwen3-vl-235b-a22b (+instruct/thinking), qwen3-vl-8b (+instruct/thinking), qwen-vl-max-2025-08-13
DeepSeek
- deepseek-v3.2
- deepseek-v3.2-thinking
- deepseek-v3-0324
Meta (Llama)
- llama-3.3-70b-instruct
- llama-4-maverick-17b-128e-instruct
Mistral
- mistral-large-3
- mistral-medium-2505 / 2508
- mistral-small-2506 / 3.1-24b-instruct-2503
- magistral-medium-2506
Other Text Models
- Baidu: ernie-5.0-preview (1103/1120), ernie-exp (various dates)
- Zhipu: glm-4.5, glm-4.5-air, glm-4.5v, glm-4.6
- MiniMax: minimax-m1, minimax-m2, minimax-m2-preview
- Tencent: hunyuan-t1-20250711, hunyuan-vision-1.5-thinking
- Amazon: nova-2-lite, amazon-nova-experimental-chat, amazon.nova-pro-v1:0
- Misc: command-a-03-2025, ling-1t, ling-flash-2.0, step-3, ring-flash-2.0, intellect-3
Image Generation Models
Google (Imagen/Gemini)
- gemini-3-pro-image-preview (Standard, 2k, and 4k versions)
- gemini-2.5-flash-image-preview
- gemini-2.0-flash-preview-image-generation
- imagen-4.0-generate-001
- imagen-4.0-fast-generate-001
- imagen-4.0-ultra-generate-001
- imagen-3.0-generate-002
Black Forest Labs (Flux)
- flux-2-pro
- flux-2-dev
- flux-2-flex
- flux-1-kontext-pro
- flux-1-kontext-dev
- flux-1-kontext-max
OpenAI
- dall-e-3
- gpt-image-1
- gpt-image-1-mini
- gpt-image-1-high-fidelity
Alibaba (Qwen Image)
- qwen-image-edit
- qwen-image-prompt-extend
Tencent (Hunyuan)
- hunyuan-image-3.0
- hunyuan-image-3.0-fal
- hunyuan-image-2.1
Wan / Video Models
- wan2.5-preview
- wan2.5-t2i-preview (Text to Image)
- wan2.5-i2i-preview (Image to Image)
- vidu-q2-image
- reve-v1 / reve-fast-edit
Other Visual Models
- recraft-v3
- ideogram-v3-quality
- seedream-3 / seedream-4.5 / seedream-4-high-res-fal
- seededit-3.0
- mai-image-1
- photon
- lucid-origin
- hazel-gen-2 / 4
- hazel-edit-2 / 6
- hidream-e1.1
- tangerine
- ghost-pepper
Hidden / Anonymous / Battle Models
These are internal codenames, blind test models, or obfuscated names specific to the Arena.
The "Beluga/Phantom" Series:
- beluga-1128-1
- phantom-1203-1
- phantom-mm-1125-1
The "Raptor" Series:
- raptor (base, 1110, 1119, 1123, 1124, 1202)
- raptor-llm (1017, 1024, 1125, 1205)
- raptor-vision (1015, 1107)
The "EB / X1" Series:
- EB45-turbo
- EB45-turbo-vl-0906
- EB45-vision
- x1-1-preview-0915
- x1-turbo-0906
Anonymous IDs:
- anonymous-1111, 1010, 915, 922, 925
- lmarena-internal-test-only
- not-a-new-model
- stephen-v2 / stephen-vision-csfix
Abstract Codenames:
- aegis-core, blackhawk, blitzphase, bridge-mind, dark-dragon, dashspark, evo-logic
- flashstride, flying-octopus, frame-flow, gauss, holo-scope, integrated-info
- leepwal, micro-mango, monster, monterey, neon, newton
- nightride-on / v2, rain-drop, redwood, route66, rushstream
- seahawk, silentnova, silvandra, skyhawk, sunshine-ai
- swiftflare, voltwhirl, viper, whisperfall, winter-wind