Qwen3 Max Thinking (Preview)
Nov 2025
Qwen3 Max Thinking (Preview)Compare inon
Benchmarks of Qwen3 Max Thinking (Preview)13
BenchmarkByScoreRank
Overall1
Knowledge4
Science2
Coding1
Documents1
Legacy · saturated or no longer reported4
No benchmark matches
—per task?
API price not publicprice
—output speed?
262Kcontext
Ranking?
Alternatives?
ModelScoreCostPer taskWhat changes
BetterGPT-6 Lunalow+49%—$0.005Function calling only with none · Chat Completions+3
OpenDeepSeek V4 Pro 0813max+80%—$0.67Pass back reasoning_content · Chat Completions+4BetterGPT-5.6 Lunalow+45%—$0.01
BetterClaude Haiku 5.5low+22%—$0.024Remove sampling parameters+5BetterGPT-6.1 Terralow+75%—$0.055ModelScoreCostPer taskWhat changes
Nothing here yet.
ModelScoreCostPer taskWhat changes
GPT-6 Lunalow+49%—$0.005Function calling only with none · Chat Completions+3GPT-5.6 Lunalow+45%—$0.01Claude Haiku 5.5low+22%—$0.024Remove sampling parameters+5GPT-6.1 Terralow+75%—$0.055
Gemini 3.1 Flash Litehigh+33%—$0.06ModelScoreCostPer taskWhat changes
DeepSeek V4 Pro 0813maxOpen+80%—$0.67Pass back reasoning_content · Chat Completions+4
Kimi K3maxOpen+74%—$2.00Sampling parameters are fixed+3DeepSeek V4.1 Flash 1017maxOpen+57%—$0.26Pass back reasoning_content · Chat Completions+4TInklingxhighOpen+52%—$0.60Context: 262K → 256K
Step 5 PreviewthinkingOpen+52%—$1.03web_search not available · Chat Completions+2ModelScoreCostPer taskWhat changes
Qwen3.6 Max Previewthinking+38%—$1.30 / $7.80Include the word JSON · QwenCloudQwen3.8 Max (0902)thinking+16%—$5.41Qwen3.8 2.4T A95BthinkingOpen+14%—$2.16max_tokens range · QwenCloud+4Qwen3.7 Maxthinking+14%—$0.93Qwen3.5 397B A17BthinkingOpen+13%—$0.47Qwen3 Max Thinkingthinking+9%—$1.20 / $6Context: 262K → 256KQwen3.6 Plusthinkingsame—$0.50 / $3Qwen3.8 FlashthinkingOpen-10%—$0.37max_tokens range · QwenCloud+5All Alibaba (Qwen) models →History
2 versionsBenchmarks
all categoriesChat & Knowledge
Humanity's Last Exam12.7#160
AA Omniscience Index-39.1#173
CritPt (physics)0#182
AA Intelligence Index 4.3.2≈16.3#140
AA Omniscience · Accuracy27.4#94
GPQA Diamond77.6#132
AIME 2025 · AA archived82.3#39
Coding
LiveCodeBench · AA archived53.5#81
Agents & Tool Use
τ²-bench · AA83.6#70
Writing & Language
IFBench (instructions)53.8#105
FAQ
How much does Qwen3 Max Thinking (Preview) cost?›
There is no per-unit API price: API price not public.
Is Qwen3 Max Thinking (Preview) free to use?›
There is no free API.
How good is Qwen3 Max Thinking (Preview), and what is it best at?›
#159 of 336 in Chat & Knowledge, #99 of 239 in Coding, #68 of 242 in Agents & Tool Use, #105 of 245 in Writing & Language.
Is Qwen3 Max Thinking (Preview) better than Kimi K2 0711?›
On AA Omniscience · Accuracy, Qwen3 Max Thinking (Preview) scores 27.4 vs 27.4 — about the same.
What is the context window of Qwen3 Max Thinking (Preview)?›
262K tokens.
When was Qwen3 Max Thinking (Preview) released? Is there a newer version?›
Released 2025-11-03. It has been superseded by Qwen3 Max Thinking.
Is Qwen3 Max Thinking (Preview) open source?›
No, Qwen3 Max Thinking (Preview) is a closed model available through an API. The closest open-weights option here is DeepSeek V4 Pro 0813.
Can Qwen3 Max Thinking (Preview) read images?›
No, it accepts text only.
What is Qwen3 Max Thinking (Preview)?›
Qwen3 Max Thinking (Preview) is a closed language model from Alibaba (Qwen), released Nov 2025.
What are the best alternatives to Qwen3 Max Thinking (Preview)?›
GPT-6 Luna (better), DeepSeek V4 Pro 0813 (open weights).