Model comparison
Kimi K2.7 Code vs Llama-3.3-70B-Instruct
Kimi K2.7 Code is the stronger model overall, scoring 43.3 to 30.6 on the Noometry Index. Llama-3.3-70B-Instruct costs 11× 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.
MetaSummary
- They share 7 benchmarks with published results for both. Kimi K2.7 Code scores higher in 4 categories and Llama-3.3-70B-Instruct in 1 category; 5 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K2.7 Code leads 52.9 to 15.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 95.6% for Kimi K2.7 Code and 5.1% for Llama-3.3-70B-Instruct.
- Llama-3.3-70B-Instruct is cheaper at $0.10 / $0.32 per million input/output tokens, against $0.95 / $4 for Kimi K2.7 Code.
- Kimi K2.7 Code accepts more context: 262K tokens versus 128K.
Side by side
| Kimi K2.7 Code | Llama-3.3-70B-Instruct | |
|---|---|---|
| Provider | Moonshot AI | Meta |
| Noometry Index | 43.3 | 30.6 |
| Released | 2026-06-12 | 2024-12-06 |
| Weights | Open | Open |
| Context window | 262K | 128K |
| Max output | 262K | 4K |
| Input $ / M tokens | $0.95 | $0.10 |
| Output $ / M tokens | $4 | $0.32 |
| Results tracked | 19 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K2.7 Code leads
Kimi K2.7 Code: 42.9 (#95), Llama-3.3-70B-Instruct: 31.0 (#290)
| Benchmark | Kimi K2.7 Code | Llama-3.3-70B-Instruct |
|---|---|---|
| SciCode | 47.5% | 26% |
| WeirdML | 54.1% | 14.4% |
| DeepSWE | 30.5% | — |
| FrontierCode | 30.1% | — |
| LMArena WebDev | 1473 | — |
| BigCodeBench Instruct | — | 46.9% |
| LiveBench Coding | — | 36.6% |
| LMArena Coding | — | 1268 |
| BigCodeBench Complete | — | 57.5% |
| ALE-Bench | 886.23 | — |
Agentic & Tool Use Llama-3.3-70B-Instruct leads
Kimi K2.7 Code: 24.0 (#122), Llama-3.3-70B-Instruct: 25.8 (#105)
| Benchmark | Kimi K2.7 Code | Llama-3.3-70B-Instruct |
|---|---|---|
| APEX-Agents | 37.6% | — |
| Berkeley Function Calling Leaderboard | — | 31.9% |
| BALROG | — | 23% |
| GBAEval | 0.9% | — |
| Vending-Bench 2 | 5,083 | — |
Reasoning Kimi K2.7 Code leads
Kimi K2.7 Code: 39.0 (#61), Llama-3.3-70B-Instruct: 14.1 (#327)
| Benchmark | Kimi K2.7 Code | Llama-3.3-70B-Instruct |
|---|---|---|
| SimpleBench | 57.9% | 19.9% |
| CritPt | 10% | 0% |
| Epoch Capabilities Index | 149.97 | 127.33 |
| Chess Puzzles | 21% | — |
| LiveBench Reasoning | — | 50.8% |
| LMArena Hard Prompts | — | 1257 |
| DTBench | — | 59.5% |
| LiveBench Data Analysis | — | 49.5% |
| LMCA | — | 17.5% |
| Surface Evolver Bench | 48.8% | — |
| ForecastBench | — | 58.6 |
| LiveBench | — | 50.2% |
Math Kimi K2.7 Code leads
Kimi K2.7 Code: 52.9 (#48), Llama-3.3-70B-Instruct: 15.3 (#298)
| Benchmark | Kimi K2.7 Code | Llama-3.3-70B-Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 95.6% | 5.1% |
| FrontierMath (Tiers 1-3) | 54% | — |
| FrontierMath Tier 4 | 12.2% | — |
| LiveBench Math | — | 42.2% |
| LMArena Math | — | 1267 |
| MATH Level 5 | — | 41.6% |
Knowledge Kimi K2.7 Code leads
Kimi K2.7 Code: 53.5 (#57), Llama-3.3-70B-Instruct: 30.6 (#226)
| Benchmark | Kimi K2.7 Code | Llama-3.3-70B-Instruct |
|---|---|---|
| GPQA Diamond | 87.9% | 47.4% |
| SimpleQA Verified | 36.5% | — |
| Confabulations | — | 22.8% |
| Vectara Hallucination Rate | — | 4.1% |
| LMArena Expert | — | 1225 |
| MMLU | — | 86.3% |
Multilingual Not comparable
Kimi K2.7 Code: —, Llama-3.3-70B-Instruct: 39.9 (#220)
| Benchmark | Kimi K2.7 Code | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Non-English | — | 1236 |
| LMArena Chinese | — | 1217 |
| LMArena French | — | 1281 |
| LMArena German | — | 1251 |
| LMArena Japanese | — | 1150 |
| LMArena Korean | — | 1143 |
| LMArena Russian | — | 1252 |
| LMArena Spanish | — | 1270 |
Instruction Following Not comparable
Kimi K2.7 Code: —, Llama-3.3-70B-Instruct: 71.1 (#157)
| Benchmark | Kimi K2.7 Code | Llama-3.3-70B-Instruct |
|---|---|---|
| LiveBench Instruction Following | — | 82.7% |
| LMArena Instruction Following | — | 1242 |
Long Context Not comparable
Kimi K2.7 Code: —, Llama-3.3-70B-Instruct: 26.4 (#295)
| Benchmark | Kimi K2.7 Code | Llama-3.3-70B-Instruct |
|---|---|---|
| Fiction.LiveBench | — | 33.3% |
| LMArena Longer Query | — | 1256 |
Writing & Preference Not comparable
Kimi K2.7 Code: —, Llama-3.3-70B-Instruct: 47.6 (#207)
| Benchmark | Kimi K2.7 Code | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Text | — | 1274 |
| LMArena Creative Writing | — | 1250 |
| LMArena Multi-Turn | — | 1280 |
| LiveBench Language | — | 39.2% |
Frequently asked questions
Is Kimi K2.7 Code better than Llama-3.3-70B-Instruct?
Kimi K2.7 Code is the stronger model overall, scoring 43.3 to 30.6 on the Noometry Index. Llama-3.3-70B-Instruct costs 11× 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, Kimi K2.7 Code or Llama-3.3-70B-Instruct?
Llama-3.3-70B-Instruct is cheaper. It lists at $0.10 per million input tokens and $0.32 per million output tokens; Kimi K2.7 Code lists at $0.95 and $4.
Is Kimi K2.7 Code or Llama-3.3-70B-Instruct better for coding?
Kimi K2.7 Code scores higher on coding benchmarks: 42.9 versus 31.0 in the Noometry coding category.
Which has the bigger context window?
Kimi K2.7 Code does, with 262K tokens against 128K.
How many benchmarks do Kimi K2.7 Code and Llama-3.3-70B-Instruct share?
7 benchmarks have published results for both models. Kimi K2.7 Code has 19 scored results on Noometry and Llama-3.3-70B-Instruct has 43.