Compare / Qwen3-Max vs Kimi K2 (0905)
Qwen3-Max vs Kimi K2 (0905)
Which is cheaper, which is more capable, and which does better at each kind of work, from 16 results on 3 sources that measured both models.
Of the 16 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.
Results as of 9 October 2026 (catalogue release 2026-10-09-dd620e778ac5). See also Alibaba vs Moonshot AI, every model from both labs.
Qwen3-Max or Kimi K2 (0905)? The short answer
They score about the same, so pick Kimi K2 (0905): it costs 31% less per token.
- Kimi K2 (0905) is cheaper: $1.07 against $1.56 per million tokens, 31% less.
- They score about the same on the Artificial Analysis Intelligence Index (15.6 and 15.3).
- Qwen3-Max is newer: released 24 Sep 2025, against 5 Sep 2025 for Kimi K2 (0905).
- On third-party benchmarks, on the results both models have: Qwen3-Max does better for reasoning and knowledge; Kimi K2 (0905) for writing.
Head to head
Every fact side by side
| Qwen3-Max | Kimi K2 (0905) | |
|---|---|---|
| Price per million tokens | $0.78 in · $3.90 out | $0.60 in · $2.50 out |
| Blended (3 in : 1 out) | $1.56 | $1.07 |
| Cheapest host, blended | – | – |
| Intelligence Index | 15.6 | 15.3 |
| Spring Prompt overall | – | – |
| Output speed, tokens a second | – | – |
| Context window | 262,144 tokens | 262,144 tokens |
| Released | 24 Sep 2025 (newer) | 5 Sep 2025 |
| Weights | Closed | 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 the models on sale today, for each kind of work, from our own benchmarks and licensed sources. Further right is better; the better rank is in green.
Every shared result
Each model's best result on every measure both have, grouped by benchmark, with the setting that produced it. Under each measure, every model that benchmark has measured is a grey tick, best to the right, so you can see where the two sit in the field. The headline measures of each benchmark come first; the rest are under "Show all". Green marks the better value; where a source reports ranges that overlap, neither is marked.
| Measure | Qwen3-Max | Kimi K2 (0905) |
|---|---|---|
| Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results | ||
| Overall · TextArena rating, higher is better 108 · 113 of 387 measured | 1,423 | 1,419 |
| Business, management and finance · TextArena rating, higher is better 110 · 134 of 380 measured | 1,419 | 1,404 |
| Creative writing · TextArena rating, higher is better 105 · 119 of 385 measured | 1,393 | 1,381 |
| Artificial Analysis reported by Artificial Analysis · 3 results | ||
| Artificial Analysis Intelligence Indexindex score, higher is better 123 · 126 of 243 measured | 15.6 | 15.3 |
| GPQA Diamond% of questions, higher is better 132 · 130 of 231 measured | 76.4% | 76.7% |
| Humanity's Last Exam% of questions, higher is better 140 · 178 of 242 measured | 11.9% | 6.4% |
| UGI Leaderboard reported by UGI Leaderboard · 3 results | ||
| Writing score · UGIscore out of 100, higher is better 115 · 61 of 230 measured | 38.6 | 54.4 |
| Requested-length error · UGI% off the requested word count, lower is better 134 · 129 of 230 measured | 23.0% | 22.0% |
| Style adherence · UGIscore from 0 to 1, higher is better 155 · 194 of 230 measured | 0.33 | 0.31 |
Show all 16 results (7 more)
| Measure | Qwen3-Max | Kimi K2 (0905) |
|---|---|---|
| Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results | ||
| Expert prompts · TextArena rating, higher is better 137 · 138 of 338 measured | 1,417 | 1,417 |
| Instruction following · TextArena rating, higher is better 106 · 136 of 387 measured | 1,413 | 1,390 |
| Writing, literature and language · TextArena rating, higher is better 101 · 133 of 386 measured | 1,405 | 1,382 |
| Artificial Analysis reported by Artificial Analysis · 4 results | ||
| IFBench% of instructions, higher is better 118 · 126 of 195 measured | 44.2% | 41.7% |
| Long-context reasoning (AA-LCR)% of questions, higher is better 154 · 146 of 233 measured | 50.0% | 53.0% |
| Terminal-Bench Hard% of tasks, higher is better 91 · 84 of 190 measured | 20.5% | 23.5% |
| τ²-bench telecom% of tasks, higher is better 79 · 80 of 193 measured | 74.3% | 73.4% |
Compare them with other models
Quick answers
Is Qwen3-Max better than Kimi K2 (0905)?
They score about the same, so pick Kimi K2 (0905): it costs 31% less per token.
Which is cheaper, Qwen3-Max or Kimi K2 (0905)?
Kimi K2 (0905) is cheaper: $1.07 against $1.56 per million tokens, 31% less.
Which is better for reasoning and knowledge, Qwen3-Max or Kimi K2 (0905)?
Qwen3-Max ranks 121 of 198 models on sale for hard questions across science, maths and general knowledge, against 128 for Kimi K2 (0905).
Which is better for writing, Qwen3-Max or Kimi K2 (0905)?
Kimi K2 (0905) ranks 73 of 188 models on sale for what people prefer in blind comparisons, and judged writing quality, against 99 for Qwen3-Max.
Which is better for sticking to the facts, Qwen3-Max or Kimi K2 (0905)?
Qwen3-Max ranks 50 of 152 models on sale for summarising without inventing things, and factual answers, against 135 for Kimi K2 (0905).
Run Qwen3-Max and Kimi K2 (0905) on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.