Compare / Qwen3 235B A22B vs DeepSeek V3.1
Qwen3 235B A22B vs DeepSeek V3.1
Which is cheaper, which is more capable, and which does better at each kind of work, from 15 results on 3 sources that measured both models.
Qwen3 235B A22B or DeepSeek V3.1? The short answer
- Qwen3 235B A22B is cheaper: $0.80 against $0.82 per million tokens, 3% less.
- DeepSeek V3.1 scores higher on the Artificial Analysis Intelligence Index: 13.7 against 9.5.
- On our business benchmarks, DeepSeek V3.1 ranks higher for reasoning and knowledge, writing and sticking to the facts.
At a glance
| Qwen3 235B A22B | DeepSeek V3.1 | |
|---|---|---|
| Price per million tokens | $0.46 in · $1.82 out | $0.55 in · $1.65 out |
| Blended (3 in : 1 out) | $0.80 | $0.82 |
| Cheapest host, blended | – | $0.42 (DeepInfra, FP4) |
| Intelligence Index | 9.5 | 13.7 |
| Spring Prompt overall | – | – |
| Output speed, tokens a second | – | – |
| Context window | 131,072 tokens | 163,840 tokens |
| Weights | Open | Open |
| Developer based in | China | China |
Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.
Which is better for what
Each model's rank among models available today, for each kind of work, from our own benchmarks and licensed sources. The better rank is in green.
| Use case | Qwen3 235B A22B | DeepSeek V3.1 |
|---|---|---|
| Reasoning and knowledgeHard questions across science, maths and general knowledge | 137 of 195 | 114 of 195 |
| WritingWhat people prefer in blind comparisons, and judged writing quality | 102 of 187 | 82 of 187 |
| Sticking to the factsSummarising without inventing things, and factual answers | 64 of 153 | 20 of 153 |
Every shared result
| Benchmark | Metric | Qwen3 235B A22B | DeepSeek V3.1 |
|---|---|---|---|
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,414 | 1,417 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,366 | 1,402 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,395 | 1,432 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,381 | 1,416 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,403 | 1,417 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,378 | 1,407 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 9.50 | 13.7 |
| Artificial Analysisreported | GPQA Diamond% of questions | 70.0% | 77.9% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 11.0% | 14.3% |
| Artificial Analysisreported | IFBench% of instructions | 38.7% | 41.5% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 0.0% | 56.7% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 6.1% | 25.0% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 27.2% | 37.4% |
| Vectara Hallucination Leaderboardreported | Answer rate% of documents | 94.9% | 94.5% |
| Vectara Hallucination Leaderboardreported | Hallucination rate% of summaries | 9.3% | 5.5% |
Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.
Quick answers
Is Qwen3 235B A22B better than DeepSeek V3.1?
It depends on the task. Among models available today, on our benchmarks and licensed sources, DeepSeek V3.1 ranks higher for reasoning and knowledge, writing and sticking to the facts.
Which is cheaper, Qwen3 235B A22B or DeepSeek V3.1?
Qwen3 235B A22B costs $0.80 per million tokens (three input to one output), against $0.82 for DeepSeek V3.1.
Run Qwen3 235B A22B and DeepSeek V3.1 on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.