OverviewAll LeaderboardsMethodology
New Chat

Text Arena

Which model is the best at everyday chat?

Nov 20, 2026
1,502,280 votes
424 models

Text Arena

Frontier
AnthropicClaude Fable 5.2 (Max)1,529 rating. 9 points above rank 2.
Daily Driver
Gemini 3.8 Flash (High)Near-frontier performance at 1/10th the price.
Efficient
DeepSeek-V4-ProSolid performance at 1/100th the price.
1,4001,500
  1. 1,529
    Anthropic
    Claude Fable 5.2Max
  2. 1,520
    Gemini 4 ArgonHigh
  3. 1,516
    Anthropic
    Claude Opus 5.5High
  4. 1,515
    Gemini 4 CarbonHigh
  5. 1,502
    Gemini 4 NeonHigh
  6. 1,502
    Anthropic
    Claude Fable 5High
  7. 1,498
    Anthropic
    Claude Opus 4.6High
  8. 1,492
    Anthropic
    Claude Fable 5.1Max
  9. 1,491
    Gemini 4 HydrogenHigh
  10. 1,490
    Anthropic
    Claude Opus 4.7High
  11. 1,489
    Gemini 3.8 FlashHigh
  12. 1,487
    Gemini 3.7 FlashHigh
  13. 1,487
    Meta
    Muse WonderMax
  14. 1,484
    Gemini 3 Pro
  15. 1,482
    Anthropic
    Claude Opus 4.7
Model RankingScoreCost per task (P50)Blended price $/M (3:1)
  1. 1AnthropicClaude Fable 5.2 (Max)1,529$0.24$32
  2. 2Gemini 4 Argon (High)1,520$0.03$8
  3. 3AnthropicClaude Opus 5.5 (High)1,516$0.08$16
  4. 4Gemini 4 Carbon (High)1,515$0.05$8
  5. 5Gemini 4 Neon (High)1,502$0.08$6
  6. 6AnthropicClaude Fable 5 (High)1,502$0.14$40
  7. 7AnthropicClaude Opus 4.6 (High)1,498$0.07$20
  8. 8AnthropicClaude Fable 5.1 (Max)1,492$0.46$40
  9. 9Gemini 4 Hydrogen (High)1,491$0.05$6
  10. 10AnthropicClaude Opus 4.7 (High)1,490$0.06$20
  11. 11Gemini 3.8 Flash (High)1,489$0.01$3
  12. 12Gemini 3.7 Flash (High)1,487$0.01$3
  13. 13MetaMuse Wonder (Max)1,487$0.10$16
  14. 14Gemini 3 Pro1,484N/A$9.50
  15. 15AnthropicClaude Opus 4.71,482$0.06$20

Pareto frontier

Best score for the cost
1,2501,3001,3501,4001,4501,500
Anthropic
$0.001$0.01$0.10

Median cost per task

Frontier
AnthropicClaude Fable 5.2 (Max)1,529 rating. 9 points above rank 2.
Daily Driver
Gemini 3.8 Flash (High)Near-frontier performance at 1/10th the price.
Efficient
DeepSeek-V4-ProSolid performance at 1/100th the price.
On the frontierScore$/task
  1. AnthropicClaude Fable 5.2 (Max)1,529$0.24
  2. Gemini 4 Argon (High)1,520$0.03
  3. Gemini 3.8 Flash (High)1,489$0.01
  4. Grok 4.20 Beta 11,464$0.01
  5. Grok 4.51,451$0.01
  6. MiMo-V2.6-Pro1,450$0.01
  7. DeepSeek-V4-Pro1,446$0.01
  8. DeepSeek-V4-Flash Preview (High)1,412$0.01
  9. DeepSeek-V4-Flash1,407$0.01
  10. Granite 4.1 8B1,266$0.01
313 models not plotted

These models are on the leaderboard but do not appear on the Pareto chart because they have insufficient cost data. The numbers beside their names show their leaderboard ranks.

  • 14Gemini 3 Pro
  • 26Claude Opus 4.5 (High, 32K, 20251101)
  • 29Muse Spark
  • 32Claude Opus 4.5 (20251101)
  • 36Gemini 3 Flash
  • 38Claude Sonnet 4.5 (20250929)
  • 47Claude Sonnet 4.5 (High, 32K, 20250929)
  • 53GLM-5
  • 57Claude Opus 4.1 (Thinking 16K, 20250805)
  • 61Gemini 2.5 Pro
  • 64Claude Opus 4.1 (20250805)
  • 69Qwen3.6-Max-Preview
  • 71GPT-4.5 Preview (2025-02-27)
  • 73Claude Opus 4 (Thinking 16K, 20250514)
  • 79Grok 4.1
  • 83GPT-5.1 (High)
  • 85Grok 4.1 (Thinking)
  • 91ChatGPT-4o Latest (20250326)
  • 92ERNIE 5.0 (0110)
  • 93ERNIE-5.0-Preview (1203)
  • 94Gemma 4 31B
  • 95Claude Opus 4 (20250514)
  • 96Grok 4.1 Fast (Reasoning)
  • 102DeepSeek-V3.2-Exp
  • 104GPT-5.1
  • 105ERNIE-5.0-Preview (1022)
  • 109DeepSeek-V3.1-Terminus
  • 110GLM-4.7
  • 112GPT-4.1 (2025-04-14)
  • 113DeepSeek-V3.1 (Thinking)
  • 115GLM-4.6
  • 116DeepSeek-V3.2
  • 117Gemma 4 26B A4B
  • 119Grok 3 Preview (02-24)
  • 120Claude Sonnet 4 (Thinking 32K, 20250514)
  • 122Gemini 2.5 Flash
  • 123Grok 4 (0709)
  • 125Qwen3-Max-Preview
  • 126Claude 3.7 Sonnet (Thinking 32K, 20250219)
  • 127Qwen3-Max (2025-09-23)
  • 128Kimi K2.5 (Instant)
  • 131Kimi K2 Thinking Turbo
  • 132DeepSeek-V3.2 (Thinking)
  • 134DeepSeek-V3.2-Exp (Thinking)
  • 135DeepSeek-V3 (0324)
  • 136LongCat-Flash-Chat Exp (2602)
  • 137NVIDIA-Nemotron-3.5-Super-VL-120B-A12B (NVFP4)
  • 138DeepSeek-R1 (0528)
  • 139GPT-5 Chat
  • 141Claude Sonnet 4 (20250514)
  • 142DeepSeek-V3.1
  • 143Beam 1.0 501B A23B
  • 145GPT-5.2 (High)
  • 146DeepSeek-V3.1-Terminus (Thinking)
  • 147NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)
  • 148Grok 4 Fast (Reasoning)
  • 149o3 (2025-04-16)
  • 150Kimi K2 Preview (0905)
  • 151Grok 4 Fast (Non-Reasoning)
  • 152o1 (2024-12-17)
  • 153Gemini 2.5 Flash Preview (09-2025)
  • 154Qwen3-235B-A22B-Instruct (2507)
  • 155Mistral Medium 3.1 (2508)
  • 156Claude 3.7 Sonnet (20250219)
  • 157Hunyuan Vision 1.5 Thinking
  • 158Hunyuan T1 (20250711)
  • 160DeepSeek-R1
  • 161Gemini 2.5 Flash-Lite Preview (Thinking, 06-17)
  • 162GLM-4.5
  • 163GPT-5 (High)
  • 164Claude 3.5 Sonnet (20241022)
  • 165Kimi K2 Preview (0711)
  • 166Qwen3-235B-A22B-Thinking (2507)
  • 167Muse Glimmer
  • 168Qwen3-235B-A22B (Non-Thinking)
  • 169o1-preview
  • 171Mistral Medium 3 (2505)
  • 173Qwen3-Coder-480B-A35B-Instruct
  • 174Hunyuan TurboS (20250416)
  • 176Gemini 2.5 Flash-Lite Preview (Non-Thinking, 09-2025)
  • 178Gemini 1.5 Pro (002)
  • 182Qwen3-VL-235B-A22B-Instruct
  • 184Qwen2.5-Max
  • 185MiMo-V2-Flash (Non-Thinking)
  • 187GPT-4.1 mini (2025-04-14)
  • 188GLM-4.6V
  • 189DeepSeek-V3
  • 190amazon-nova-experimental-chat (26-01-10)
  • 191Gemma 3 27B IT
  • 192MiniMax-M2.1 Preview
  • 193Gemini Advanced (0514)
  • 194Gemini 2.0 Flash-Lite Preview (02-05)
  • 196Gemini 2.0 Flash (001)
  • 198amazon-nova-experimental-chat (26-02-10)
  • 200amazon-nova-experimental-chat (12-10)
  • 201Qwen3-VL-235B-A22B-Thinking
  • 203GPT-4o (2024-05-13)
  • 204o4-mini (2025-04-16)
  • 206Command A (03-2025)
  • 207Grok 3 Mini Beta
  • 208MiMo-V2-Flash (Thinking)
  • 209Step-2 16K Exp (202412)
  • 211Gemma 3 12B IT
  • 212Grok 3 Mini (High)
  • 213Llama 3.1 Nemotron Ultra 253B v1
  • 214LongCat-Flash-Chat
  • 215GLM-4.5-Air
  • 216GPT-4o (2024-08-06)
  • 217Qwen3-Next-80B-A3B-Thinking
  • 218Gemini 1.5 Pro (001)
  • 219GPT-4 Turbo (2024-04-09)
  • 220Mistral Small 3.2 (2506)
  • 221GPT-5 mini (High)
  • 222Qwen3-235B-A22B
  • 224MiniMax-M1
  • 225Qwen3-30B-A3B-Instruct (2507)
  • 226Grok 2 (2024-08-13)
  • 227Qwen-Plus (0125)
  • 228GLM-4-Plus (0111)
  • 229Qwen3-Next-80B-A3B-Instruct
  • 231o3-mini (High)
  • 232Step-3
  • 233GLM-4.5V
  • 234DeepSeek-V2.5 (1210)
  • 235INTELLECT-3
  • 236GPT-4.1 nano (2025-04-14)
  • 237Llama 4 Maverick 17B 128E Instruct
  • 238Claude 3.5 Haiku (20241022)
  • 239GLM-4.7-Flash
  • 240Llama 3.1 405B Instruct (FP8)
  • 241Claude 3.5 Sonnet (20240620)
  • 242Llama 3.3 Nemotron Super 49B v1.5
  • 243Qwen3-32B
  • 244Hunyuan Large (2025-02-10)
  • 245Hunyuan TurboS (20250226)
  • 246Hunyuan Turbo (0110)
  • 247NVIDIA-Nemotron-3-Super-120B-A12B
  • 248Llama 3.1 405B Instruct (BF16)
  • 249o3-mini
  • 250amazon-nova-experimental-chat (11-10)
  • 251Gemma 3n E4B IT
  • 252Gemini 1.5 Flash (002)
  • 253Llama 3.3 Nemotron Super 49B v1
  • 254Yi Lightning
  • 255GPT-4o mini (2024-07-18)
  • 256Olmo 3.1 32B Instruct
  • 257QwQ-32B
  • 258Gemma 2 27B IT
  • 259Step-1o Turbo (202506)
  • 260Mercury 2
  • 261GLM-4-Plus
  • 262Llama 4 Scout 17B 16E Instruct
  • 263Claude 3 Opus (20240229)
  • 264Mistral Large 2 (2407)
  • 265amazon-nova-experimental-chat (10-09)
  • 266GPT-4 Turbo Preview (1106)
  • 267Llama 3.3 70B Instruct
  • 268Qwen-Max (0919)
  • 269Gemma 2 9B IT SimPO
  • 270Qwen3-30B-A3B
  • 271NVIDIA-Nemotron-3.5-Lightning-30B-A3B (NVFP4)
  • 272MiniMax-M2
  • 273o1-mini
  • 274GPT-4 Turbo Preview (0125)
  • 275Magistral Medium (2506)
  • 276gpt-oss-120b
  • 277Hunyuan Large Vision
  • 278amazon-nova-experimental-chat (10-20)
  • 279Hunyuan Standard (2025-02-10)
  • 280Mistral Large (2411)
  • 281Llama 3.1 Nemotron 70B Instruct
  • 282Gemma 3 4B IT
  • 283Qwen2.5-Plus (1127)
  • 284Grok 2 mini (2024-08-13)
  • 285Athene 70B (0725)
  • 286Mistral Small 3.1 24B Instruct (2503)
  • 287Amazon Nova 2 Lite
  • 288Ling-flash-2.0
  • 289Reka Core (20240904)
  • 290GPT-4 (0613)
  • 292DeepSeek-V2.5
  • 293Granite 4.2 30B
  • 294Olmo 3 32B Think
  • 295Gemini 1.5 Flash (001)
  • 296Command R+ (08-2024)
  • 297Llama 3.1 Nemotron 51B Instruct
  • 298Jamba 1.5 Large
  • 299Gemma 2 9B IT
  • 300Llama 3.1 70B Instruct
  • 301Ring-flash-2.0
  • 302GPT-4 (0314)
  • 303Llama 3 70B Instruct
  • 304Qwen2.5-72B-Instruct
  • 305Athene V2 Chat
  • 306Olmo 3.1 32B Think
  • 307GPT-5 nano (High)
  • 308Llama 3.1 Tulu 3 70B
  • 309Claude 3 Sonnet (20240229)
  • 310Reka Flash (20240904)
  • 311Granite-4.0-H-Small
  • 312NVIDIA-Nemotron-3-Nano-30B-A3B (BF16)
  • 313GLM-4 (0520)
  • 314gpt-oss-20b
  • 315Nemotron-4 340B Instruct
  • 316Gemini 1.5 Flash-8B (001)
  • 317Amazon Nova Pro 1.0
  • 318Command R+
  • 319Aya Expanse 32B
  • 320Mistral Small 24B Instruct (2501)
  • 321OLMo 2 32B Instruct (0325)
  • 322Qwen2-72B-Instruct
  • 323Ministral 8B (2410)
  • 324Amazon Nova Lite 1.0
  • 325Mercury
  • 326Claude 3 Haiku (20240307)
  • 327Jamba 1.5 Mini
  • 328Mistral Large (2402)
  • 329Phi-4
  • 330Command R (08-2024)
  • 331Qwen2.5-Coder-32B-Instruct
  • 332DeepSeek-Coder-V2
  • 333Gemini Pro (Dev API)
  • 334Granite 4.2 3B
  • 335Granite 4.2 8B
  • 336WizardLM 70B
  • 337Amazon Nova Micro 1.0
  • 338Llama 3 8B Instruct
  • 339Llama 3.1 Tulu 3 8B
  • 340Command R
  • 341Hunyuan Standard 256K
  • 342Qwen1.5-110B-Chat
  • 343Mistral Medium
  • 344GPT-3.5 Turbo (0125)
  • 345Mixtral 8x22B Instruct v0.1
  • 346Aya Expanse 8B
  • 347Qwen1.5-72B-Chat
  • 348Gemini Pro
  • 349OpenChat 3.5
  • 350Gemma 2 2B IT
  • 351Reka Flash 21B (Online, 20240226)
  • 352Llama 3.1 8B Instruct
  • 353Reka Flash 21B (20240226)
  • 354Vicuna 33B
  • 355Granite 3.1 8B Instruct
  • 356Zephyr ORPO 141B A35B v0.1
  • 357DBRX Instruct Preview
  • 358Mixtral 8x7B Instruct v0.1
  • 359Yi 1.5 34B Chat
  • 360OpenHermes 2.5 Mistral 7B
  • 361GPT-3.5 Turbo (1106)
  • 362Yi 34B Chat
  • 363Falcon 180B Chat
  • 364Gemma 1.1 7B IT
  • 365Phi-3 Medium 4K Instruct
  • 366OpenChat 3.5 (0106)
  • 367Granite 3.1 2B Instruct
  • 368Tulu 2 DPO 70B
  • 369Llama 3.2 3B Instruct
  • 370Snowflake Arctic Instruct
  • 371WizardLM 13B
  • 372SOLAR 10.7B Instruct v1.0
  • 373Qwen1.5-14B-Chat
  • 374Granite 3.0 8B Instruct
  • 375Starling LM 7B Alpha
  • 376DeepSeek LLM 67B Chat
  • 377Zephyr 7B Beta
  • 378Guanaco 33B
  • 379Dolphin 2.2.1 Mistral 7B
  • 380Nous Hermes 2 Mixtral 8x7B DPO
  • 381Phi-3 Small 8K Instruct
  • 382MPT 30B Chat
  • 383Qwen1.5-32B-Chat
  • 384InternLM2.5 20B Chat
  • 385Vicuna 13B
  • 386Llama 2 70B SteerLM Chat
  • 387Llama 2 70B Chat
  • 388Starling LM 7B Beta
  • 389Mistral 7B Instruct v0.2
  • 390QwQ-32B-Preview
  • 391Qwen-14B-Chat
  • 392Gemma 7B IT
  • 393Zephyr 7B Alpha
  • 394Granite 3.0 2B Instruct
  • 395Phi-3 Mini 4K Instruct (June 2024)
  • 396Gemma 1.1 2B IT
  • 397Alpaca 13B
  • 398Llama 2 13B Chat
  • 399Mistral 7B Instruct
  • 400Vicuna 7B
  • 401StripedHyena Nous 7B
  • 402PaLM 2
  • 403Code Llama 34B Instruct
  • 404Phi-3 Mini 128K Instruct
  • 405Llama 3.2 1B Instruct
  • 406Qwen1.5-7B-Chat
  • 407SmolLM2 1.7B Instruct
  • 408Gemma 2B IT
  • 409Phi-3 Mini 4K Instruct
  • 410Llama 2 7B Chat
  • 411GPT4All 13B Snoozy
  • 412MPT 7B Chat
  • 413Qwen1.5-4B-Chat
  • 414Koala 13B
  • 415ChatGLM3 6B
  • 416OpenAssistant Pythia 12B
  • 417RWKV-4 Raven 14B
  • 418ChatGLM2 6B
  • 419OLMo 7B Instruct
  • 420FastChat-T5 3B
  • 421Dolly 2.0 12B
  • 422ChatGLM 6B
  • 423LLaMA 13B
  • 424StableLM Tuned Alpha 7B

Default Leaderboard Plots

Previous pillar · Agent
Tool Use Index

Which model is the best at using tools?

Next
Search

Which model is the best at online search?

USE CASES

  • Chat with AI
  • Build Apps & Websites
  • Write & Edit Text
  • Search the Web
  • Generate Images
  • Generate Videos
  • Chose any model
  • Compare Models Side by Side

LEADERBOARD RANKINGS

  • Overall
  • Agent
  • Alignment Index
  • Text
  • WebDev
  • Image-to-WebDev
  • Text to Image
  • Image Edit
  • Text to Video
  • Image to Video
  • Video Edit
  • Vision
  • Document
  • Search

COMPANY

  • About Us
  • How It Works
  • Blog
  • Careers
  • Leaderboard Changelog
  • Product Changelog
  • Help Center
  • FAQ

LEGAL

  • Terms
  • Privacy
  • Cookies

FOLLOW

  • X
  • LinkedIn
  • YouTube
  • Discord

© Arena Intelligence