Model comparison
DeepSeek-V3 vs Kimi K2.7 Code
Kimi K2.7 Code is the stronger model overall, scoring 43.0 to 39.2 on the Noometry Index. DeepSeek-V3 costs 4.2× less per token, which makes it the better buy when Kimi K2.7 Code's lead doesn't matter for your workload.
Last verified . 7 shared benchmarks.
DeepSeekSummary
- They share 7 benchmarks with published results for both. DeepSeek-V3 scores higher in 0 categories and Kimi K2.7 Code in 4 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K2.7 Code leads 52.6 to 30.4.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 37.8% for DeepSeek-V3 and 95.6% for Kimi K2.7 Code.
- DeepSeek-V3 is cheaper at $0.24 / $0.90 per million input/output tokens, against $0.95 / $4 for Kimi K2.7 Code.
- Kimi K2.7 Code accepts more context: 262K tokens versus 164K.
Side by side
| DeepSeek-V3 | Kimi K2.7 Code | |
|---|---|---|
| Provider | DeepSeek | Moonshot AI |
| Noometry Index | 39.2 | 43.0 |
| Released | 2024-12-26 | 2026-06-12 |
| Weights | Open | Open |
| Context window | 164K | 262K |
| Max output | 164K | 262K |
| Input $ / M tokens | $0.24 | $0.95 |
| Output $ / M tokens | $0.90 | $4 |
| Results tracked | 60 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
DeepSeek-V3: 41.9 (#99), Kimi K2.7 Code: 42.4 (#92)
| Benchmark | DeepSeek-V3 | Kimi K2.7 Code |
|---|---|---|
| SciCode | 35.8% | 47.5% |
| WeirdML | 36.1% | 54.1% |
| DeepSWE | — | 30.5% |
| FrontierCode | — | 30.1% |
| Aider Polyglot | 55.1% | — |
| LMArena WebDev | — | 1473 |
| BigCodeBench Instruct | 50% | — |
| LiveBench Coding | 70.9% | — |
| LMArena Coding | 1368 | — |
| BigCodeBench Complete | 62.2% | — |
| ALE-Bench | — | 886.23 |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73% | — |
Agentic & Tool Use Not comparable
DeepSeek-V3: —, Kimi K2.7 Code: 23.9 (#115)
| Benchmark | DeepSeek-V3 | Kimi K2.7 Code |
|---|---|---|
| APEX-Agents | — | 37.6% |
| GBAEval | — | 0.9% |
| METR Time Horizons | 49.6% | — |
| Vending-Bench 2 | — | 5,083 |
Reasoning Kimi K2.7 Code leads
DeepSeek-V3: 20.0 (#226), Kimi K2.7 Code: 38.2 (#56)
| Benchmark | DeepSeek-V3 | Kimi K2.7 Code |
|---|---|---|
| SimpleBench | 27.2% | 57.9% |
| CritPt | 0% | 10% |
| Epoch Capabilities Index | 135.95 | 150.01 |
| Kagi LLM Benchmark | 52.3% | — |
| Chess Puzzles | — | 21% |
| LiveBench Reasoning | 65.8% | — |
| LMArena Hard Prompts | 1365 | — |
| DTBench | 64.8% | — |
| LiveBench Data Analysis | 60.9% | — |
| LMCA | 15.5% | — |
| Surface Evolver Bench | — | 48.8% |
| BIG-Bench Hard | 87.5% | — |
| ForecastBench | 59.1 | — |
| HellaSwag | 88.9% | — |
| LiveBench | 66.9% | — |
| PIQA | 84.7% | — |
| WinoGrande | 85.2% | — |
Math Kimi K2.7 Code leads
DeepSeek-V3: 30.4 (#208), Kimi K2.7 Code: 52.6 (#39)
| Benchmark | DeepSeek-V3 | Kimi K2.7 Code |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 95.6% |
| FrontierMath (Tiers 1-3) | — | 54% |
| FrontierMath Tier 4 | — | 12.2% |
| Omni-MATH | 40.3% | — |
| LiveBench Math | 73.5% | — |
| LMArena Math | 1373 | — |
| MATH Level 5 | 75.5% | — |
| FrontierMath (Feb 2025 set) | 1.7% | — |
Knowledge Kimi K2.7 Code leads
DeepSeek-V3: 37.5 (#147), Kimi K2.7 Code: 53.6 (#51)
| Benchmark | DeepSeek-V3 | Kimi K2.7 Code |
|---|---|---|
| GPQA Diamond | 67.6% | 87.9% |
| SimpleQA Verified | — | 36.5% |
| MMLU-Pro | 72.3% | — |
| Confabulations | 26.1% | — |
| Vectara Hallucination Rate | 6.1% | — |
| GPQA (HELM) | 53.8% | — |
| LMArena Expert | 1352 | — |
| ARC (AI2) Challenge | 95.3% | — |
| MMLU | 87.2% | — |
| TriviaQA | 82.9% | — |
Multilingual Not comparable
DeepSeek-V3: 48.8 (#138), Kimi K2.7 Code: —
| Benchmark | DeepSeek-V3 | Kimi K2.7 Code |
|---|---|---|
| LMArena Non-English | 1358 | — |
| LMArena Chinese | 1390 | — |
| LMArena French | 1386 | — |
| LMArena German | 1374 | — |
| LMArena Japanese | 1333 | — |
| LMArena Korean | 1321 | — |
| LMArena Russian | 1372 | — |
| LMArena Spanish | 1358 | — |
Instruction Following Not comparable
DeepSeek-V3: 73.0 (#125), Kimi K2.7 Code: —
| Benchmark | DeepSeek-V3 | Kimi K2.7 Code |
|---|---|---|
| LiveBench Instruction Following | 81.5% | — |
| IFEval | 83.2% | — |
| LMArena Instruction Following | 1345 | — |
Long Context Not comparable
DeepSeek-V3: 34.0 (#248), Kimi K2.7 Code: —
| Benchmark | DeepSeek-V3 | Kimi K2.7 Code |
|---|---|---|
| Fiction.LiveBench | 50% | — |
| LMArena Longer Query | 1353 | — |
Writing & Preference Not comparable
DeepSeek-V3: 57.6 (#124), Kimi K2.7 Code: —
| Benchmark | DeepSeek-V3 | Kimi K2.7 Code |
|---|---|---|
| LMArena Text | 1375 | — |
| LMArena Creative Writing | 1364 | — |
| Short-Story Creative Writing | 77% | — |
| EQ-Bench Creative Writing | 1472 | — |
| WildBench | 83% | — |
| LMArena Multi-Turn | 1389 | — |
| LiveBench Language | 49.1% | — |
Frequently asked questions
Is DeepSeek-V3 better than Kimi K2.7 Code?
Kimi K2.7 Code is the stronger model overall, scoring 43.0 to 39.2 on the Noometry Index. DeepSeek-V3 costs 4.2× less per token, which makes it the better buy when Kimi K2.7 Code's lead doesn't matter for your workload.
Which is cheaper, DeepSeek-V3 or Kimi K2.7 Code?
DeepSeek-V3 is cheaper. It lists at $0.24 per million input tokens and $0.90 per million output tokens; Kimi K2.7 Code lists at $0.95 and $4.
Is DeepSeek-V3 or Kimi K2.7 Code better for coding?
They score almost the same on coding (41.9 vs 42.4); test both on your own repository before choosing.
Which has the bigger context window?
Kimi K2.7 Code does, with 262K tokens against 164K.
How many benchmarks do DeepSeek-V3 and Kimi K2.7 Code share?
7 benchmarks have published results for both models. DeepSeek-V3 has 60 scored results on Noometry and Kimi K2.7 Code has 19.