Model specification · Qwen

Qwen3 235B A22B Thinking 2507

qwen/qwen3-235b-a22b-thinking-2507

Qwen3-235B-A22B-Thinking-2507 is a high-performance, open-weight Mixture-of-Experts (MoE) language model optimized for complex reasoning tasks. It activates 22B of its 235B parameters per forward pass and natively supports up to 262,144...

SpecificationsReported values
Context window
262K
Maximum output
Architecture
text->text
Tokenizer
Qwen3
Knowledge cutoff
2025-06-30
Moderated
No
Input modalities
text
Output modalities
text
Full pricingUSD where reported
PriceStandard
Input$0.230 / M
Output$2.30 / M
Benchmark indexesReported indexes
Intelligence
19.6
Coding
22.1
Agentic
3.8
Design Arena Elo
1027
Example workload costsStandard pricing only
WorkloadInput / outputCache shareEstimated cost
Quick check10K / 1K × 10%$0.00460
Long analysis100K / 4K × 150%$0.032
Workday20K / 2K × 1,00030%$9.20
More from QwenAll Qwen models
ModelContextInput / MOutput / MIntel.
Qwen3.7 Max1M$1.48$4.4246
Qwen3.6 Plus1M$0.325$1.9539.6
Qwen3.7 Plus1M$0.320$1.2839
Qwen3.6 27B262K$0.289$2.4037.1
Qwen3.5 397B A17B262K$0.390$2.3433.7
Qwen3.5-122B-A10B262K$0.260$2.0832.3
Qwen3.6 35B A3B262K$0.140$1.0031.6
Qwen3.5-35B-A3B262K$0.140$1.0024
Qwen3.5-9B262K$0.100$0.15021.4
Qwen3 Coder Next262K$0.120$0.80021.1
Qwen3 Next 80B A3B Thinking262K$0.150$1.2016.7
Qwen3 30B A3B Thinking 250782K$0.200$2.4014.4