Compare / Qwen3-235B-A22B-Instruct-2507 vs Kimi K2 (0711)

Qwen3-235B-A22B-Instruct-2507 vs Kimi K2 (0711)

Which is cheaper, which is more capable, and which does better at each kind of work, from 24 results on 4 sources that measured both models.

AlibabaQwen3-235B-A22B-Instruct-2507 Released 21 Jul 2025 · $0.15 in · $0.75 out
Moonshot AIKimi K2 (0711) Released 12 Jul 2025 · $0.57 in · $2.30 out
Qwen3-235B-A22B-Instruct-2507 wins 103 too close to callKimi K2 (0711) wins 11

Of the 24 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.

Qwen3-235B-A22B-Instruct-2507 or Kimi K2 (0711)? The short answer

They score about the same, so pick Qwen3-235B-A22B-Instruct-2507: it costs 3.3 times less per token.

  • Qwen3-235B-A22B-Instruct-2507 is cheaper: $0.30 against $1.00 per million tokens, 3.3 times less.
  • They score about the same on the Artificial Analysis Intelligence Index (12.0 and 12.7).
  • Qwen3-235B-A22B-Instruct-2507 is newer: released 21 Jul 2025, against 12 Jul 2025 for Kimi K2 (0711).
  • On third-party benchmarks, on the results both models have: Kimi K2 (0711) does better for agents and tool use, reasoning and knowledge and writing.

Head to head

12.0
Intelligence Index
12.7
$0.30
Price per million tokens, blendedlower is better
$1.00
262,144
Context window, tokens
131,072

Every fact side by side

Qwen3-235B-A22B-Instruct-2507Kimi K2 (0711)
Price per million tokens$0.15 in · $0.75 out$0.57 in · $2.30 out
Blended (3 in : 1 out)$0.30$1.00
Cheapest host, blended$0.15 (GMICloud, FP8)–
Intelligence Index12.012.7
Spring Prompt overall––
Output speed, tokens a second––
Context window262,144 tokens131,072 tokens
Released21 Jul 2025 (newer)12 Jul 2025
WeightsOpenOpen
Developer based inChinaChina

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.

Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls 51 of 164 19 of 164
Reasoning and knowledgeHard questions across science, maths and general knowledge 129 of 198 132 of 198
WritingWhat people prefer in blind comparisons, and judged writing quality 85 of 188 82 of 188
Sticking to the factsSummarising without inventing things, and factual answers 98 of 152 not measured

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.

MeasureQwen3-235B-A22B-Instruct-2507Kimi K2 (0711)
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Overall · TextArena rating, higher is better 109 · 114 of 387 measured 1,422 1,418
Business, management and finance · TextArena rating, higher is better 97 · 127 of 380 measured 1,432 1,409
Creative writing · TextArena rating, higher is better 122 · 131 of 385 measured 1,378 1,371
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 153 · 147 of 243 measured 12.0 12.7
GPQA Diamond% of questions, higher is better 136 · 131 of 231 measured 75.3% 76.6%
Humanity's Last Exam% of questions, higher is better 144 · 168 of 242 measured 11.1% 7.4%
Berkeley Function Calling Leaderboard (BFCL) V4 reported by Berkeley Function Calling Leaderboard (BFCL) V4 · 3 results
Overall accuracy · BFCL% correct, higher is better 20 · 9 of 81 measured 52.2%bfcl prompt 59.1%bfcl fc
Irrelevance detection · BFCL% correct, higher is better 41 · 17 of 81 measured 81.7%bfcl fc 87.3%bfcl fc
Memory · BFCL% correct, higher is better 26 · 16 of 81 measured 23.9%bfcl fc 29.0%bfcl fc
UGI Leaderboard reported by UGI Leaderboard · 3 results
Writing score · UGIscore out of 100, higher is better 80 · 71 of 230 measured 49.0 52.5
Requested-length error · UGI% off the requested word count, lower is better 69 · 129 of 230 measured 15.0% 22.0%
Style adherence · UGIscore from 0 to 1, higher is better 64 · 217 of 230 measured 0.36 0.29
Show all 24 results (12 more)
MeasureQwen3-235B-A22B-Instruct-2507Kimi K2 (0711)
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Expert prompts · TextArena rating, higher is better 103 · 141 of 338 measured 1,447 1,416
Instruction following · TextArena rating, higher is better 104 · 151 of 387 measured 1,414 1,381
Writing, literature and language · TextArena rating, higher is better 115 · 141 of 386 measured 1,396 1,375
Artificial Analysis reported by Artificial Analysis · 4 results
IFBench% of instructions, higher is better 113 · 127 of 195 measured 46.1% 41.5%
Long-context reasoning (AA-LCR)% of questions, higher is better 177 · 146 of 233 measured 33.9% 53.0%
Terminal-Bench Hard% of tasks, higher is better 110 · 107 of 190 measured 15.2% 15.9%
τ²-bench telecom% of tasks, higher is better 127 · 95 of 193 measured 33.3% 61.1%
Berkeley Function Calling Leaderboard (BFCL) V4 reported by Berkeley Function Calling Leaderboard (BFCL) V4 · 5 results
Multi-turn tasks · BFCL% correct, higher is better 20 · 15 of 81 measured 45.4%bfcl fc 50.6%bfcl fc
Relevance detection · BFCL% correct, higher is better 8 · 49 of 81 measured 93.8%bfcl prompt 75.0%bfcl fc
Single-turn calls (curated) · BFCL% correct, higher is better 2 · 55 of 81 measured 90.3%bfcl prompt 81.6%bfcl fc
Single-turn calls (user-contributed) · BFCL% correct, higher is better 15 · 15 of 81 measured 78.7%bfcl prompt 78.7%bfcl fc
Web search · BFCL% correct, higher is better 20 · 16 of 81 measured 54.0%bfcl fc 66.5%bfcl fc

Quick answers

Is Qwen3-235B-A22B-Instruct-2507 better than Kimi K2 (0711)?

They score about the same, so pick Qwen3-235B-A22B-Instruct-2507: it costs 3.3 times less per token.

Which is cheaper, Qwen3-235B-A22B-Instruct-2507 or Kimi K2 (0711)?

Qwen3-235B-A22B-Instruct-2507 is cheaper: $0.30 against $1.00 per million tokens, 3.3 times less.

Which is better for agents and tool use, Qwen3-235B-A22B-Instruct-2507 or Kimi K2 (0711)?

Kimi K2 (0711) ranks 19 of 164 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 51 for Qwen3-235B-A22B-Instruct-2507.

Which is better for reasoning and knowledge, Qwen3-235B-A22B-Instruct-2507 or Kimi K2 (0711)?

Qwen3-235B-A22B-Instruct-2507 ranks 129 of 198 models on sale for hard questions across science, maths and general knowledge, against 132 for Kimi K2 (0711).

Which is better for writing, Qwen3-235B-A22B-Instruct-2507 or Kimi K2 (0711)?

Kimi K2 (0711) ranks 82 of 188 models on sale for what people prefer in blind comparisons, and judged writing quality, against 85 for Qwen3-235B-A22B-Instruct-2507.

Which has the bigger context window, Qwen3-235B-A22B-Instruct-2507 or Kimi K2 (0711)?

Qwen3-235B-A22B-Instruct-2507: 262,144 tokens, against 131,072 for Kimi K2 (0711).

Run Qwen3-235B-A22B-Instruct-2507 and Kimi K2 (0711) on your own prompt

Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.

Try the playgroundTalk to us