RSS

Lab · China

Alibaba

AI models from Alibaba: 164 models in the catalog, version timeline, prices and sourced benchmarks.

  • 164 models in the catalog
  • 125 with open weights
  • 39 closed or without public weights
  • 49 with sourced benchmarks
  • 62 with a page here

We count grouped models: Instruct, Base, quantized and size variants of the same line become one. Behind that there are 496 Hugging Face repositories, 44 OpenRouter endpoints and 130 historical records at Epoch AI.

Version timeline by family

Main model lines

202420242025202520262026QwenQwenQwen, 20232.0Qwen 2.0, 2023Qwen 1, 20231.5Qwen 1.5, 20242Qwen 2, 20242.5Qwen 2.5, 20243Qwen 3, 20253.5Qwen 3.5, 2026Qwen 3.6, 20263.7Qwen 3.7, 2026Qwen 3.8, 2026Qwen-AudioQwen-AudioQwen-Audio, 2023Qwen-VL-MaxQwen-VL-MaxQwen-VL-Max, 2024Qwen PlusQwen PlusQwen Plus, 2024CodeQwen1.5CodeQwen 1.5, 2024QwQQwQQwQ, 2024QWQ-PlusQWQ-Plus, 2025Wan2.1Wan 2.1, 20252.2Wan 2.2, 2025Wan 2.5, 2025Qwen-ImageQwen-Image-2512Qwen-Image-2512, 20252.1Qwen-Image 2.1, 2026Qwen-Image-EditQwen-Image-Edit-2511Qwen-Image-Edit-2511, 2025202420242025202520262026QwenQwenQwen, 2023Qwen 2.0, 2023Qwen 1, 20231.5Qwen 1.5, 20242Qwen 2, 2024Qwen 2.5, 20243Qwen 3, 20253.5Qwen 3.5, 2026Qwen 3.6, 2026Qwen 3.7, 20263.8Qwen 3.8, 2026Qwen-AudioQwen-Audio, 2023Qwen-VL-MaxQwen-VL-Max, 2024Qwen PlusQwen Plus, 2024CodeQwen1.5CodeQwen 1.5, 2024QwQQwQQwQ, 2024QWQ-Plus, 2025Wan2.1Wan 2.1, 20252.2Wan 2.2, 2025Wan 2.5, 2025Qwen-ImageQwen-Image-2512, 20252.1Qwen-Image 2.1, 2026Qwen-Image-EditQwen-Image-Edit-2511, 2025
Research and others (25)
2022202220232023202420242025202520262026M6-TM6-TM6-T, 2021PLUGPLUGPLUG, 2021AliceMind-MMUAliceMind-M…AliceMind-MMU, 2021AR-LDMAR-LDMAR-LDM, 2022ONE-PEACEONE-PEACEONE-PEACE, 2023mPLUG-Owl2mPLUG-Owl 2, 2023SeaLLM-7B-v2SeaLLM-7B-v2SeaLLM-7B-v2, 2023SeaLLM-7B-v2.5, 2023RecGPT / Xingchen LLM…RecGPT / Xingchen LLM (淘宝（中国）软件有限公司)RecGPT / Xi…RecGPT / Xingchen LLM (淘宝（中国）软件有限公司), 2024Ovis2Ovis 2, 2024Ovis, 20241.6Ovis 1.6, 20242.5Ovis 2.5, 2025ChunkLlama2ChunkLlama 2, 2024Animate AnyoneAnimate Any…Animate Anyone, 20242Animate Anyone 2, 2025OutfitAnyoneOutfitAnyoneOutfitAnyone, 2024Alibaba-NLP (mGTE)Alibaba-NLP…Alibaba-NLP (mGTE), 2024Marco-o1Marco-o 1, 2024XunguangXunguangXunguang, 2024QVQQVQQVQ, 2024QVQ-Max, 2025GTE ModernbertGTE Modernb…GTE Modernbert, 2025AgentFounderAgentFounderAgentFounder, 2025Tongyi DeepResearchTongyi Deep…Tongyi DeepResearch, 2025Qwen-Image-LayeredQwen-Image-…Qwen-Image-Layered, 2025WebWorldWebWorldWebWorld, 2026Qwen-Image-BenchQwen-Image-…Qwen-Image-Bench, 2026Qwen-AgentWorldQwen-AgentW…Qwen-AgentWorld, 2026Wan-DancerWan-DancerWan-Dancer, 2026Qwen-Drive1.0Qwen-Drive 1.0, 20262023202320252025M6-TM6-TM6-T, 2021PLUGPLUGPLUG, 2021AliceMind-MMUAliceMind-MMU, 2021AR-LDMAR-LDMAR-LDM, 2022ONE-PEACEONE-PEACE, 2023mPLUG-Owl2mPLUG-Owl 2, 2023SeaLLM-7B-v2SeaLLM-7B-v2, 2023SeaLLM-7B-v2.5, 2023RecGPT / Xingchen LLM (淘宝（中国）软件有限…RecGPT / Xingchen LLM (淘宝（中国）软件有限公司)RecGPT / Xingchen LLM (淘宝（中国）软件有限公司), 2024Ovis2Ovis 2, 2024Ovis, 2024Ovis 1.6, 20242.5Ovis 2.5, 2025ChunkLlama2ChunkLlama 2, 2024Animate AnyoneAnimate Anyone, 20242Animate Anyone 2, 2025OutfitAnyoneOutfitAnyone, 2024Alibaba-NLP (mGTE)Alibaba-NLP (mGTE), 2024Marco-o1Marco-o 1, 2024XunguangXunguang, 2024QVQQVQQVQ, 2024QVQ-Max, 2025GTE ModernbertGTE Modernbert, 2025AgentFounderAgentFounder, 2025Tongyi DeepResearchTongyi DeepResearch, 2025Qwen-Image-LayeredQwen-Image-Layered, 2025WebWorldWebWorld, 2026Qwen-Image-BenchQwen-Image-Bench, 2026Qwen-AgentWorldQwen-AgentWorld, 2026Wan-DancerWan-Dancer, 2026Qwen-Drive1.0Qwen-Drive 1.0, 2026

This group holds benchmarks, datasets, research artifacts, and lines from brands that are not in this area's catalog.

Tap a point to see the version.Tap or hover over a point to see the version. Dates come from the sources: OpenRouter listing or Hugging Face repo creation, which may differ from the official launch. Family and version were grouped by rule from the name.

Version scores over time

0%25%50%75%100%20252026GPQA Diamond: Qwen1.5-72B, 28.8%, 2024GPQA Diamond: Qwen1.5-32B, 30.7%, 2024GPQA Diamond: Qwen2-72B, 40.8%, 2024GPQA Diamond: Qwen2.5-32B, 46.1%, 2024GPQA Diamond: Qwen2.5-72B, 49.1%, 2024GPQA Diamond: Qwen2.5-7B, 35.5%, 2024GPQA Diamond: Qwen-Turbo, 41.8%, 2024GPQA Diamond: Qwen-Plus, 48.1%, 2025GPQA Diamond: Qwen2.5-Max, 56.1%, 2025GPQA Diamond: QwQ-32B, 65.3%, 2025GPQA Diamond: QWQ-Plus, 65.4%, 2025GPQA Diamond: Qwen3-8B, 56.8%, 2025GPQA Diamond: Qwen3-235B-A22B, 80.1%, 2025GPQA Diamond: Qwen3 14B, 63.8%, 2025GPQA Diamond: Qwen3 30B A3B, 70.1%, 2025GPQA Diamond: Qwen3 32B, 65.7%, 2025GPQA Diamond: Qwen3-1.7B, 38%, 2025GPQA Diamond: Qwen3-4B, 52.3%, 2025GPQA Diamond: Qwen3-Max, 72.6%, 2025GPQA Diamond: Qwen3.5 397B A17B, 86.4%, 2026GPQA Diamond: Qwen 3.5 Plus (hosted 397B-A17B), 84.8%, 2026GPQA Diamond: Qwen3.5-35B-A3B, 83.5%, 2026GPQA Diamond: Qwen3.5-9B, 79%, 2026GPQA Diamond: Qwen3.5-Flash, 82.3%, 2026GPQA Diamond: Qwen3.6 Plus, 88.4%, 2026GPQA Diamond: Qwen3.6 35B A3B, 84.8%, 2026GPQA Diamond: Qwen 3.6 Max (Preview), 87.4%, 2026GPQA Diamond: Qwen3.6 27B, 85.9%, 2026GPQA Diamond: Qwen3.6 Flash, 83.3%, 2026GPQA Diamond: Qwen3.7 Max, 90.9%, 2026GPQA Diamond: Qwen3.7 Plus, 87.9%, 2026GPQA Diamond: Qwen3.7 Flash, 82.3%, 2026GPQA Diamond: Qwen3.8 Max, 92.7%, 2026GPQA Diamond (OpenRouter): Qwen2.5 72B Instruct, 47.7%, 2024GPQA Diamond (OpenRouter): Qwen2.5 VL 72B Instruct, 45.7%, 2025GPQA Diamond (OpenRouter): Qwen3-235B-A22B, 77.4%, 2025GPQA Diamond (OpenRouter): Qwen3 14B, 61.2%, 2025GPQA Diamond (OpenRouter): Qwen3 30B A3B, 60.9%, 2025GPQA Diamond (OpenRouter): Qwen3 32B, 63.3%, 2025GPQA Diamond (OpenRouter): Qwen3-Coder, 61.2%, 2025GPQA Diamond (OpenRouter): Qwen3 235B A22B Thinking 2507, 79.1%, 2025GPQA Diamond (OpenRouter): Qwen3 30B A3B Instruct 2507, 64.8%, 2025GPQA Diamond (OpenRouter): Qwen3 Coder 30B A3B Instruct, 53.4%, 2025GPQA Diamond (OpenRouter): Qwen3 Next 80B A3B Instruct, 71.8%, 2025GPQA Diamond (OpenRouter): Qwen3 VL 235B A22B Instruct, 71.1%, 2025GPQA Diamond (OpenRouter): Qwen3 VL 30B A3B Instruct, 66.8%, 2025GPQA Diamond (OpenRouter): Qwen3 VL 8B Instruct, 53.9%, 2025GPQA Diamond (OpenRouter): Qwen3 Coder Next, 76.3%, 2026GPQA Diamond (OpenRouter): Qwen3.5 397B A17B, 88.1%, 2026GPQA Diamond (OpenRouter): Qwen3.5-35B-A3B, 82.9%, 2026GPQA Diamond (OpenRouter): Qwen3.5-9B, 81%, 2026GPQA Diamond (OpenRouter): Qwen3.5-122B-A10B, 84.8%, 2026GPQA Diamond (OpenRouter): Qwen3.6 35B A3B, 85.1%, 2026GPQA Diamond (OpenRouter): Qwen3.6 27B, 85.1%, 2026GPQA Diamond (OpenRouter): Qwen3.7 Max, 90.9%, 2026GPQA Diamond (OpenRouter): Qwen3.7 Plus, 87.9%, 2026GPQA Diamond (OpenRouter): Qwen3.8 2.4T A95B, 91.8%, 2026GPQA Diamond (OpenRouter): Qwen3.8 27B, 87.2%, 2026GPQA Diamond (OpenRouter): Qwen3.8 Flash, 90.5%, 2026OTIS Mock AIME: Qwen2.5-32B, 7.4%, 2024OTIS Mock AIME: Qwen2.5-72B, 8.1%, 2024OTIS Mock AIME: Qwen2.5-7B, 2.5%, 2024OTIS Mock AIME: Qwen-Turbo, 6.1%, 2024OTIS Mock AIME: Qwen-Plus, 17.8%, 2025OTIS Mock AIME: Qwen2.5-Max, 16.1%, 2025OTIS Mock AIME: QwQ-32B, 59.2%, 2025OTIS Mock AIME: Qwen3-8B, 56.1%, 2025OTIS Mock AIME: Qwen3-235B-A22B, 86.7%, 2025OTIS Mock AIME: Qwen3 14B, 66.4%, 2025OTIS Mock AIME: Qwen3 30B A3B, 70.3%, 2025OTIS Mock AIME: Qwen3 32B, 66.9%, 2025OTIS Mock AIME: Qwen3-1.7B, 8.1%, 2025OTIS Mock AIME: Qwen3-Max, 73.3%, 2025OTIS Mock AIME: Qwen3.5 397B A17B, 88.9%, 2026OTIS Mock AIME: Qwen 3.5 Plus (hosted 397B-A17B), 86.7%, 2026OTIS Mock AIME: Qwen3.5-35B-A3B, 70%, 2026OTIS Mock AIME: Qwen3.5-9B, 61.7%, 2026OTIS Mock AIME: Qwen3.5-4B, 55.8%, 2026OTIS Mock AIME: Qwen3.5-Flash, 84.4%, 2026OTIS Mock AIME: Qwen3.6 Plus, 93.3%, 2026OTIS Mock AIME: Qwen3.6 35B A3B, 86.7%, 2026OTIS Mock AIME: Qwen 3.6 Max (Preview), 91.1%, 2026OTIS Mock AIME: Qwen3.6 27B, 91.1%, 2026OTIS Mock AIME: Qwen3.6 Flash, 84.4%, 2026OTIS Mock AIME: Qwen3.7 Max, 95.6%, 2026OTIS Mock AIME: Qwen3.7 Plus, 93.3%, 2026OTIS Mock AIME: Qwen3.7 Flash, 86.7%, 2026OTIS Mock AIME: Qwen3.8 Max, 100%, 2026
  • GPQA Diamond
  • GPQA Diamond (OpenRouter)
  • OTIS Mock AIME

Source: Epoch AI, OpenRouter, latest version we checked.

Version prices over time (US$ per 1M tokens, log scale) ?what it is

$0.01$0.10$1.00$10.00$100.0020252026input: Qwen2.5 72B Instruct, US$ 0.36, 2024input: Qwen2.5 Coder 32B Instruct, US$ 0.66, 2024input: Qwen2.5 VL 72B Instruct, US$ 0.80, 2025input: Qwen-Plus, US$ 0.26, 2025input: Qwen3 14B, US$ 0.10, 2025input: Qwen3 30B A3B, US$ 0.12, 2025input: Qwen3 32B, US$ 0.08, 2025input: Qwen3-235B-A22B, US$ 0.14, 2025input: Qwen3-Coder, US$ 0.30, 2025input: Qwen3 235B A22B Thinking 2507, US$ 0.45, 2025input: Qwen3 30B A3B Instruct 2507, US$ 0.10, 2025input: Qwen3 Coder 30B A3B Instruct, US$ 0.15, 2025input: Qwen3 Next 80B A3B Instruct, US$ 0.09, 2025input: Qwen3 Coder Flash, US$ 0.20, 2025input: Qwen3 VL 235B A22B Instruct, US$ 0.21, 2025input: Qwen3 VL 30B A3B Instruct, US$ 0.15, 2025input: Qwen3 VL 30B A3B Thinking, US$ 0.29, 2025input: Qwen3 VL 8B Instruct, US$ 0.25, 2025input: Qwen3 Coder Next, US$ 0.12, 2026input: Qwen3.5 Plus 2026-02-15, US$ 0.26, 2026input: Qwen3.5 397B A17B, US$ 0.60, 2026input: Qwen3.5-122B-A10B, US$ 0.26, 2026input: Qwen3.5-27B, US$ 0.26, 2026input: Qwen3.5-Flash, US$ 0.065, 2026input: Qwen3.5-35B-A3B, US$ 0.14, 2026input: Qwen3.5-9B, US$ 0.052, 2026input: Qwen3.6 Plus, US$ 0.33, 2026input: Qwen3.5 Plus 2026-04-20, US$ 0.30, 2026input: Qwen3.6 Flash, US$ 0.19, 2026input: Qwen3.6 27B, US$ 0.32, 2026input: Qwen3.6 35B A3B, US$ 0.10, 2026input: Qwen3.7 Max, US$ 1.48, 2026input: Qwen3.7 Plus, US$ 0.32, 2026input: Qwen3.7 Flash, US$ 0.03, 2026input: Qwen3.8 2.4T A95B, US$ 2.00, 2026input: Qwen3.8 27B, US$ 0.15, 2026input: Qwen3.8 Flash, US$ 0.15, 2026input: Qwen3.8 Max, US$ 2.00, 2026input: Qwen3.8 Omni Flash, US$ 0.15, 2026input: Qwen3.8 Max Prime, US$ 4.00, 2026output: Qwen2.5 72B Instruct, US$ 0.40, 2024output: Qwen2.5 Coder 32B Instruct, US$ 1.00, 2024output: Qwen2.5 VL 72B Instruct, US$ 1.00, 2025output: Qwen-Plus, US$ 0.78, 2025output: Qwen3 14B, US$ 0.22, 2025output: Qwen3 30B A3B, US$ 0.50, 2025output: Qwen3 32B, US$ 0.28, 2025output: Qwen3-235B-A22B, US$ 0.80, 2025output: Qwen3-Coder, US$ 1.00, 2025output: Qwen3 235B A22B Thinking 2507, US$ 3.50, 2025output: Qwen3 30B A3B Instruct 2507, US$ 0.30, 2025output: Qwen3 Coder 30B A3B Instruct, US$ 0.60, 2025output: Qwen3 Next 80B A3B Instruct, US$ 1.10, 2025output: Qwen3 Coder Flash, US$ 0.97, 2025output: Qwen3 VL 235B A22B Instruct, US$ 1.90, 2025output: Qwen3 VL 30B A3B Instruct, US$ 0.60, 2025output: Qwen3 VL 30B A3B Thinking, US$ 1.00, 2025output: Qwen3 VL 8B Instruct, US$ 0.75, 2025output: Qwen3 Coder Next, US$ 0.80, 2026output: Qwen3.5 Plus 2026-02-15, US$ 1.56, 2026output: Qwen3.5 397B A17B, US$ 3.60, 2026output: Qwen3.5-122B-A10B, US$ 2.08, 2026output: Qwen3.5-27B, US$ 2.60, 2026output: Qwen3.5-Flash, US$ 0.26, 2026output: Qwen3.5-35B-A3B, US$ 1.00, 2026output: Qwen3.5-9B, US$ 0.077, 2026output: Qwen3.6 Plus, US$ 1.95, 2026output: Qwen3.5 Plus 2026-04-20, US$ 1.80, 2026output: Qwen3.6 Flash, US$ 1.12, 2026output: Qwen3.6 27B, US$ 3.20, 2026output: Qwen3.6 35B A3B, US$ 0.90, 2026output: Qwen3.7 Max, US$ 4.42, 2026output: Qwen3.7 Plus, US$ 1.28, 2026output: Qwen3.7 Flash, US$ 0.13, 2026output: Qwen3.8 2.4T A95B, US$ 6.00, 2026output: Qwen3.8 27B, US$ 1.88, 2026output: Qwen3.8 Flash, US$ 0.47, 2026output: Qwen3.8 Max, US$ 6.00, 2026output: Qwen3.8 Omni Flash, US$ 0.47, 2026output: Qwen3.8 Max Prime, US$ 12.00, 2026
  • input
  • output

Source: OpenRouter, latest version we checked. Dates come from the sources: OpenRouter listing or Hugging Face repo creation, which may differ from the official launch.

Models with a page (62)

ModelFamilyWeightsUse casesInput US$/1MContextGPQADates in sources
Qwen3.8 Max PrimeQwen Max PrimeundisclosedText, AgentsUS$ 4.001,000,000no data2026
Qwen3.8 Omni FlashQwen Omni FlashundisclosedText, AgentsUS$ 0.151,000,000no data2026
Qwen-Image 2.1Qwen-ImageopenImageno datano datano data2026
Qwen3.8 FlashQwen FlashopenText, AgentsUS$ 0.151,000,00090.5%2026
Qwen3.8 27BQwen3.8openText, AgentsUS$ 0.151,048,57587.2%2026
Qwen3.8 2.4T A95BQwenopenText, AgentsUS$ 2.001,048,57691.8%2026
Qwen3.8 MaxQwen MaxclosedText, AgentsUS$ 2.001,000,00092.7%2026
Qwen3.7 FlashQwen FlashclosedText, AgentsUS$ 0.031,000,00082.3%2026
Qwen3.7 PlusQwen PlusclosedText, AgentsUS$ 0.321,000,00087.9%2026
Qwen3.7 MaxQwen MaxclosedText, Agents, CodeUS$ 1.481,000,00090.9%2026
Qwen3.5 Plus 2026-04-20Qwen PlusundisclosedText, AgentsUS$ 0.301,000,000no data2026
Qwen3.6 FlashQwen FlashundisclosedText, AgentsUS$ 0.191,000,00083.3%2026
Qwen3.6 27BQwenopenText, AgentsUS$ 0.32262,14485.9%2026
Qwen 3.6 Max (Preview)Qwen MaxundisclosedCodeno datano data87.4%2026
Qwen3.6 35B A3BQwenopenText, AgentsUS$ 0.10262,14484.8%2026
Qwen3.6 PlusQwen PlusclosedText, Agents, CodeUS$ 0.331,000,00088.4%2026
Qwen3.5-122B-A10BQwenopenText, AgentsUS$ 0.26262,14484.8%2026
Qwen3.5-27BQwenopenText, AgentsUS$ 0.26262,144no data2026
Qwen3.5-FlashQwen FlashundisclosedText, AgentsUS$ 0.0651,000,00082.3%2026
Qwen3.5-35B-A3BQwenopenText, AgentsUS$ 0.14262,14483.5%2026
Qwen3.5-9BQwenopenText, AgentsUS$ 0.052262,14479%2026
Qwen3.5-4BQwenopenTextno datano datano data2026
Qwen3.5 Plus 2026-02-15Qwen PlusundisclosedText, AgentsUS$ 0.261,000,000no data2026
Qwen 3.5 Plus (hosted 397B-A17B)Qwen PlusundisclosedTextno datano data84.8%2026
Qwen3.5 397B A17BQwenopenText, AgentsUS$ 0.60262,14486.4%2026
Qwen3 Coder NextQwen Coder NextopenText, Agents, CodeUS$ 0.12262,14476.3%2026
Qwen3-ASRQwen ASRopenTranscriptionno datano datano data2026
Qwen3-TTSQwen TTSopenVoiceno datano datano data2026
Qwen3 VL 8B InstructQwen VL InstructopenText, AgentsUS$ 0.25262,14453.9%2025
Qwen3 VL 30B A3B InstructQwen VL InstructopenText, AgentsUS$ 0.15262,14466.8%2025
Qwen3 VL 30B A3B ThinkingQwen VL ThinkingopenText, AgentsUS$ 0.29262,144no data2025
Qwen3-MaxQwen MaxclosedTextno datano data72.6%2025
Qwen3 VL 235B A22B InstructQwen VL InstructopenText, AgentsUS$ 0.21262,14471.1%2025
Qwen3 Coder FlashQwen Coder FlashundisclosedText, Agents, CodeUS$ 0.201,000,000no data2025
Qwen3 Next 80B A3B InstructQwen Next InstructopenText, AgentsUS$ 0.09262,14471.8%2025
Qwen3 Coder 30B A3B InstructQwen Coder InstructopenText, Agents, CodeUS$ 0.15262,14453.4%2025
Qwen3 30B A3B Instruct 2507Qwen InstructopenText, AgentsUS$ 0.10262,14464.8%2025
Qwen3 235B A22B Thinking 2507Qwen ThinkingopenText, AgentsUS$ 0.45128,00079.1%2025
Qwen3-CoderQwen CoderopenText, Agents, CodeUS$ 0.30262,14461.2%2025
Wan2.2WanopenVideono datano datano data2025
Qwen3-235B-A22BQwen3openText, AgentsUS$ 0.14262,14480.1%2025
Qwen3 14BQwenopenText, AgentsUS$ 0.1040,96063.8%2025
Qwen3 30B A3BQwenopenText, AgentsUS$ 0.1240,96070.1%2025
Qwen3 32BQwenopenText, AgentsUS$ 0.08131,07265.7%2025
Qwen3-1.7BQwenopenTextno datano data38%2025
Qwen3-4BQwenopenTextno datano data52.3%2025
Qwen3-8BQwenopenTextno datano data56.8%2025
QWQ-PlusQWQ PlusclosedTextno datano data65.4%2025
QwQ-32BQwQopenTextno datano data65.3%2025
Qwen2.5 VL 72B InstructQwen VL InstructopenTextUS$ 0.80128,00045.7%2025
Qwen-PlusQwen PlusclosedText, AgentsUS$ 0.261,000,00048.1%2025
Qwen2.5-MaxQwen MaxclosedTextno datano data56.1%2025
Qwen2.5 Coder 32B InstructQwen Coder InstructopenText, CodeUS$ 0.6632,768no data2024
Qwen-TurboQwen TurboclosedTextno datano data41.8%2024
Qwen2.5 7B InstructQwen InstructopenText, Agentsno datano datano data2024
Qwen2.5 72B InstructQwen InstructopenText, AgentsUS$ 0.3632,76847.7%2024
Qwen2.5-72BQwenopenTextno datano data49.1%2024
Qwen2.5-7BQwenopenTextno datano data35.5%2024
Qwen2.5-32BQwenopenTextno datano data46.1%2024
Qwen2-72BQwenopenTextno datano data40.8%2024
Qwen1.5-32BQwenopenTextno datano data30.7%2024
Qwen1.5-72BQwenopenTextno datano data28.8%2024

Other lab models in the catalog (76)

ModelModalityWeightsContextInput US$/1MDownloads (30 days)1ª
Qwen3multimodal, textoopen262,144US$ 0.0885,684,7742025
Qwen2.5multimodal, textoopen128,000US$ 0.1041,575,6662024
Qwen3.5multimodal, textoopen262,144US$ 0.1428,577,2312026
Qwen3.8multimodalopen1,048,575US$ 0.1512,805,4622026
Qwen3-Embeddingembeddings, sem rótulo, textoopenno datano data14,157,2032025
Qwen3.6multimodalopen262,144US$ 0.1011,860,9772026
Qwen2.5-Codercódigoopen32,768US$ 0.669,087,0522024
Qwen3-TTS-12Hzsem rótulo, vozopenno datano data4,583,8032026
Qwen3.8-Flash-Nextmultimodalopenno datano data3,551,4742026
Qwen3-Rerankerembeddings, textoopenno datano data4,066,2432025
Qwen2multimodal, textoopenno datano data3,461,5712024
Qwen3-ASRtranscriçãoopenno datano data3,172,6882026
Qwen3-Codercódigo, textoopen262,144US$ 0.152,862,9422025
Qwen3-TTS-12Hz-1.7B-CustomVoicevozopenno datano data1,350,4392026
Qwen-Image-Editimagemopenno datano data910,7262025
Qwen3-VL-Embeddingembeddingsopenno datano data1,177,3682026
Qwen3-Coder-Nextcódigo, textoopen262,144US$ 0.12838,0642026
Wan2.1-T2Vvídeoopenno datano data648,6892025
Qwen3-Omnimultimodalopenno datano data688,0082025
Qwen1.5textoopenno datano data532,2812024
Qwen-Image-2.1imagemopenno datano data300,0862026
Qwen3-TTS-12Hz-0.6B-CustomVoicevozopenno datano data737,9902026
Qwen2.5-Math-PRMtextoopenno datano data734,9082025
Qwen-Imageimagem, multimodalopenno datano data373,6892025
Qwen3-TTS-12Hz-1.7B-VoiceDesignvozopenno datano data607,5212026
Qwen2.5-Omnimultimodalopenno datano data378,3592025
QwQtextoopenno datano data75,7322024
Qwen3-ForcedAlignertexto, transcriçãoopenno datano data506,0412026
Qwenmultimodal, textoopenno datano data130,6392023
Qwen3-Nexttextoopen262,144US$ 0.09230,7472025
Wan2.2-Animateimagem, vídeoopenno datano data245,4622025
Qwen3Guard-Gentextoopenno datano data340,2242025
Qwen2.5-Mathtextoopenno datano data316,7142024
Wan2.1-I2V-14B-480Pvídeoopenno datano data304,9702025
Wan2.1-VACEvídeoopenno datano data234,7962025
Wan2.2-I2Vvídeoopenno datano data181,8582025
Wan2.2-TI2Vvídeoopenno datano data188,9432025
Wan2.2-T2Vvídeoopenno datano data176,6682025
Wan2.1-I2V-14B-720Pvídeoopenno datano data175,0282025
Qwen2-Audiomultimodalopenno datano data153,0002024
Qwen-Image-Layeredimagemopenno datano data94,6562025
Qwen3.8-2.4Ttextoopen1,048,576US$ 2.0035,7952026
Qwen3-VL-Rerankerembeddingsopenno datano data78,5922026
Qwen1.5-MoEtextoopenno datano data76,3502024
Wan2.2-S2Vvídeoopenno datano data53,3122025
Qwen-AgentWorldtextoopenno datano data10,0712026
Qwen3-TTS-Tokenizer-12Hzvozopenno datano data66,1642026
CodeQwen1.5códigoopenno datano data7,5612024
Wan2.2-Animate-2imagem, sem rótuloopenno datano data23,4532026
QVQmultimodalopenno datano data1,7602024
Qwen2.5-Math-RMtextoopenno datano data41,1842024
Wan2.1-FLF2V-14B-720Pimagemopenno datano data17,4462025
Qwen-Image-Benchmultimodalopenno datano data22,2732026
Qwen-Drive-1.0multimodalopenno datano data7,8712026
Qwen3-Omni-30B-A3B-Captionermultimodalopenno datano data6,8282025
Qwen2-Mathtextoopenno datano data6,0242024
Qwen-Audiomultimodal, textoopenno datano data2,8432023
WebWorldtextoopenno datano data6,5462026
Qwen3Guard-Streamembeddingsopenno datano data6,2232025
Qwen-Image-2.1-PE-I2Isem rótuloopenno datano data13,2412026
Wan-Dancervídeoopenno datano data1,0712026
Qwen-1textoopenno datano data3,5182023
qwen-plustextoclosed or without public weights1,000,000US$ 0.26no data2024
qwen3-coder-flashcódigoclosed or without public weights1,000,000US$ 0.20no data2025
qwen3.5-plus-02-15multimodalclosed or without public weights1,000,000US$ 0.26no data2026
qwen3.5-flash-02-23multimodalclosed or without public weights1,000,000US$ 0.065no data2026
qwen3.6-plusmultimodalclosed or without public weights1,000,000US$ 0.33no data2026
qwen3.5-plusmultimodalclosed or without public weights1,000,000US$ 0.30no data2026
qwen3.6-flashmultimodalclosed or without public weights1,000,000US$ 0.19no data2026
qwen3.7-maxtextoclosed or without public weights1,000,000US$ 1.48no data2026
qwen3.7-plusmultimodalclosed or without public weights1,000,000US$ 0.23no data2026
qwen3.8-maxmultimodalclosed or without public weights1,000,000US$ 1.15no data2026
qwen3.7-flashmultimodal, textoclosed or without public weights1,000,000US$ 0.03no data2026
qwen3.8-flashmultimodalopen1,000,000US$ 0.15no data2026
qwen3.8-omni-flashmultimodalclosed or without public weights1,000,000US$ 0.15no data2026
qwen3.8-max-primemultimodalclosed or without public weights1,000,000US$ 4.00no data2026

Sources

  • Hugging Face Hub, metadados públicos do Hub; licença de cada modelo no repositório, latest version we checked.
  • OpenRouter, dados públicos da API; exports de benchmarks e rankings em CC BY 4.0, latest version we checked.
  • Epoch AI, CC BY 4.0, latest version we checked.

Dates come from the sources: OpenRouter listing or Hugging Face repo creation, which may differ from the official launch. Family and version were grouped by rule from the name.