Across 6 comparable benchmarks, Kimi K2 has better reported results on 5 and worse on 1.
Moderate evidence6 of 6 shared results countedHow we compare
Kimi K2 is no longer available through its official API. API prices for this version are not shown.
Price basis: Listed rates vs listed rates. The price section below shows the applicable periods.
| Fact | QwQ-32B | Kimi K2 |
|---|---|---|
| Context window | 128K tokens | 131K tokens |
| Input / 1M | — | — |
| Output / 1M | — | — |
| Blended (4:1) | — | — |
| Status | Available | Retired |
| Released | March 6, 2025 | July 11, 2025 |
Blended is 4:1 input to output with 90% cache hits at listed rates — set your own mix.
Mean of 4 shared benchmarks grouped by capability on a 0–100 scale. How averages are built
Kimi K2 ahead on 1 of 1 counted
Split 1–1 of 2 counted
Kimi K2 ahead on 1 of 1 counted
Kimi K2 ahead on 1 of 1 counted
No fixed-scale average — see the rows in Benchmark results
No shared comparable results for Abstract reasoning, Computer & tool use, Research & search, Long context, Multimodal, Cybersecurity.
Published scores, with matching modes where available. Gaps show QwQ-32B relative to Kimi K2. Reading the results
| Benchmark | QwQ-32B | Kimi K2 | Gap |
|---|---|---|---|
| Aider-PolyglotAgentic Development · Accuracy | 20.9Thinking · no tools | 59.1≠Normal · no tools | −38.2 pts#38 vs #17 of 47 |
| AIME 2024Mathematics · Accuracy | 79.5Normal · no tools | 69.6Normal · no tools | +9.9 pts#25 vs #34 of 54 |
| MATH-500Mathematics · Accuracy | 91.0Normal · no tools | 97.4Normal · no tools | −6.4 pts#41 vs #20 of 49 |
| GPQA DiamondScientific Reasoning · Accuracy | 58.0Normal · no tools | 75.1Normal · no tools | −17.1 pts#168 vs #122 of 200 |
| Creative WritingWriting | 1258Normal · no tools | 1666Normal · no tools | 0.76×#90 vs #39 of 121 |
| MMLU-ProKnowledge Exams · Accuracy | 76.0Normal · no tools | 81.1Normal · no tools | −5.1 pts#79 vs #55 of 153 |
Agentic Development · Accuracy
Scientific Reasoning · Accuracy
Writing
Knowledge Exams · Accuracy
A dash means no result is available.
| Benchmark | QwQ-32B | Kimi K2 |
|---|---|---|
| AIME2025 | — | 54.0 |
| FrontierMath | — | 2.1 |
| FrontierMath - Tier 4 | — | 0.01 |
| HLE | — | 4.7 |
| IMO-ProofBench Advanced | — | 7.1 |
| MMLU | — | 89.5 |
| APEX-SWE | — | 17.4 |
| HumanEval | 19.0 | — |
| IOI 2024 (Vals v1) | — | 0.5 |
| IOI 2025 (Vals v1) | — | 2 |
| LiveCodeBench | — | 53.7 |
| SWE-bench Verified | — | 51.8 |
| Terminal Bench Hard | — | 15.9 |
| Terminal-Bench 1.0 | — | 37.5 |
| ECI | 138 | — |
| LiveBench | — | 48.1 |
| Fiction.liveBench | — | 61.1 |
| τ²-Bench | — | 64.3 |
| τ²-Bench - Telecom | — | 65.8 |
| ARC-AGI-1 | — | 13.3 |
| SimpleBench | — | 26.3 |
| SimpleQA | — | 31.0 |
Time-of-day providers use peak-hour rates; fixed-price providers use standard rates. Off-peak discounts remain in the published details.
Drag the handle above the bar for cache hits, the one below it for the input : output mix · 5% steps
Blended price = effective input rate at 90% cache hits × 80% + output rate × 20%
A blended cost needs a published input and output price on both sides, and one of these two does not have them yet.
QwQ-32B: Listed text rates; pricing tier not confirmed.
Kimi K2: Listed text rates; pricing tier not confirmed.
| Price / 1M tokens | QwQ-32B | Kimi K2 | Ratio |
|---|---|---|---|
| Input | — | — | — |
| Output | — | — | — |
| Cached input | — | — | — |
| Cache write | — | — | — |
| Blended | — | — | — |
Estimates use the selected token mix. Cache writes, tool fees and long-context surcharges are excluded.
Lower prices are highlighted. The cache hit rate applies to input tokens only; models without a published cached-input rate use their normal input price for those tokens.| Attribute | QwQ-32B | Kimi K2 |
|---|---|---|
| Availability | Available | Retired |
| Provider | Alibaba | Moonshot AI |
| Model type | Reasoning model | Chat model |
| First release | March 6, 2025 | July 11, 2025 |
| Context window | 128K tokens | 131K tokens |
| Max output | — | 131K tokens |
| Total parameters | 32.5B | 1.0T |
| Active parameters | — | 32B |
| Architecture | Dense | Mixture of Experts (MoE) |
| Documented capabilities | Reasoning, Multilingual | Multilingual |
| Input modalities | Text | Text |
| Output modalities | Text | Text |
| Weights | Open weights | Open weights |
| Licence | Apache 2.0 | Modified MIT |
| Commercial use | Commercial use allowed | Commercial use allowed |
These are reported-score comparisons, not controlled head-to-head tests.
What people found comparing these models, or using either one. Comments also appear on the models' own pages.
Used QwQ-32B or Kimi K2? Sign in with Google to share what worked, what did not, and which API or setup you tested.
No community comments on either model yet.
Alternatives with the most benchmark overlap for this matchup.