# Kimi K2 (Jul 2025) vs Llama-3.3-70B-Instruct

> Kimi K2 (Jul 2025) is the stronger model overall, scoring 40.9 to 30.3 on the Noometry Index. Llama-3.3-70B-Instruct costs 6.5× less per token, which makes it the better buy when Kimi K2 (Jul 2025)'s lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/kimi-k2-vs-llama-3-3-70b-instruct
- Last updated: 2026-10-10
- Shared benchmarks: 25

## Summary

- They share 25 benchmarks with published results for both. Kimi K2 (Jul 2025) scores higher in 9 categories and Llama-3.3-70B-Instruct in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K2 (Jul 2025) leads 40.1 to 14.3.
- The biggest single-benchmark swing is Fiction.LiveBench: 66.7% for Kimi K2 (Jul 2025) and 33.3% for Llama-3.3-70B-Instruct.
- Llama-3.3-70B-Instruct is cheaper at $0.10 / $0.32 per million input/output tokens, against $0.57 / $2.30 for Kimi K2 (Jul 2025).
- Kimi K2 (Jul 2025) accepts more context: 262K tokens versus 128K.

## Snapshot

| | Kimi K2 (Jul 2025) | Llama-3.3-70B-Instruct |
|---|---|---|
| Provider | Moonshot AI | Meta |
| Noometry Index | 40.9 | 30.3 |
| Rank | 134 | 283 |
| Context | 262K | 128K |
| Input $/M | $0.57 | $0.10 |
| Output $/M | $2.30 | $0.32 |
| Weights | Open | Open |

## Coding

- Kimi K2 (Jul 2025): 43.1 (#83)
- Llama-3.3-70B-Instruct: 30.6 (#284)

| Benchmark | Kimi K2 (Jul 2025) | Llama-3.3-70B-Instruct |
|---|---|---|
| WeirdML | 42.8% | 14.4% |
| LMArena Coding | 1399 | 1268 |
| SWE-bench Verified (bash only) | 63.4% | — |
| Aider Polyglot | 59.1% | — |
| SciCode | — | 26% |
| GSO | 4.9% | — |
| BigCodeBench Instruct | — | 46.9% |
| LiveBench Coding | — | 36.6% |
| BigCodeBench Complete | — | 57.5% |
| ALE-Bench | 597.5 | — |

## Agentic & Tool Use

- Kimi K2 (Jul 2025): 32.2 (#59)
- Llama-3.3-70B-Instruct: 25.6 (#100)

| Benchmark | Kimi K2 (Jul 2025) | Llama-3.3-70B-Instruct |
|---|---|---|
| Berkeley Function Calling Leaderboard | 59.1% | 31.9% |
| Terminal-Bench | 35.7% | — |
| BALROG | — | 23% |
| METR Time Horizons | 59.2% | — |

## Reasoning

- Kimi K2 (Jul 2025): 22.7 (#173)
- Llama-3.3-70B-Instruct: 13.7 (#325)

| Benchmark | Kimi K2 (Jul 2025) | Llama-3.3-70B-Instruct |
|---|---|---|
| SimpleBench | 26.3% | 19.9% |
| LMArena Hard Prompts | 1384 | 1257 |
| Epoch Capabilities Index | 146.02 | 127.32 |
| ForecastBench | 60.2 | 58.6 |
| Kagi LLM Benchmark | 64.4% | — |
| CritPt | — | 0% |
| LiveBench Reasoning | — | 50.8% |
| DTBench | — | 59.5% |
| LiveBench Data Analysis | — | 49.5% |
| LMCA | — | 17.5% |
| LiveBench | — | 50.2% |

## Math

- Kimi K2 (Jul 2025): 40.1 (#81)
- Llama-3.3-70B-Instruct: 14.3 (#292)

| Benchmark | Kimi K2 (Jul 2025) | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Math | 1396 | 1268 |
| OTIS Mock AIME 2024-2025 | — | 5.1% |
| Omni-MATH | 65.4% | — |
| LiveBench Math | — | 42.2% |
| MATH Level 5 | — | 41.6% |
| FrontierMath (Feb 2025 set) | 21.4% | — |
| FrontierMath Tier 4 (v1) | 0% | — |

## Knowledge

- Kimi K2 (Jul 2025): 36.9 (#153)
- Llama-3.3-70B-Instruct: 30.5 (#220)

| Benchmark | Kimi K2 (Jul 2025) | Llama-3.3-70B-Instruct |
|---|---|---|
| Confabulations | 20.4% | 22.8% |
| Vectara Hallucination Rate | 17.9% | 4.1% |
| LMArena Expert | 1363 | 1225 |
| GPQA Diamond | — | 47.4% |
| MMLU-Pro | 81.9% | — |
| GPQA (HELM) | 65.3% | — |
| MMLU | — | 86.3% |

## Multilingual

- Kimi K2 (Jul 2025): 49.8 (#125)
- Llama-3.3-70B-Instruct: 40.1 (#215)

| Benchmark | Kimi K2 (Jul 2025) | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Non-English | 1372 | 1236 |
| LMArena Chinese | 1416 | 1217 |
| LMArena French | 1379 | 1280 |
| LMArena German | 1386 | 1251 |
| LMArena Japanese | 1349 | 1149 |
| LMArena Korean | 1324 | 1142 |
| LMArena Russian | 1384 | 1252 |
| LMArena Spanish | 1383 | 1269 |

## Instruction Following

- Kimi K2 (Jul 2025): 71.3 (#151)
- Llama-3.3-70B-Instruct: 71.3 (#152)

| Benchmark | Kimi K2 (Jul 2025) | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Instruction Following | 1347 | 1242 |
| LiveBench Instruction Following | — | 82.7% |
| IFEval | 85% | — |

## Long Context

- Kimi K2 (Jul 2025): 41.2 (#140)
- Llama-3.3-70B-Instruct: 26.4 (#290)

| Benchmark | Kimi K2 (Jul 2025) | Llama-3.3-70B-Instruct |
|---|---|---|
| Fiction.LiveBench | 66.7% | 33.3% |
| LMArena Longer Query | 1353 | 1256 |
| CL-bench | 17.6% | — |

## Writing & Preference

- Kimi K2 (Jul 2025): 62.5 (#76)
- Llama-3.3-70B-Instruct: 47.8 (#203)

| Benchmark | Kimi K2 (Jul 2025) | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Text | 1380 | 1274 |
| LMArena Creative Writing | 1349 | 1250 |
| LMArena Multi-Turn | 1371 | 1279 |
| Short-Story Creative Writing | 85.6% | — |
| EQ-Bench Creative Writing | 1666 | — |
| WildBench | 86.2% | — |
| LiveBench Language | — | 39.2% |

## FAQ

### Is Kimi K2 (Jul 2025) better than Llama-3.3-70B-Instruct?

Kimi K2 (Jul 2025) is the stronger model overall, scoring 40.9 to 30.3 on the Noometry Index. Llama-3.3-70B-Instruct costs 6.5× less per token, which makes it the better buy when Kimi K2 (Jul 2025)'s lead doesn't matter for your workload.

### Which is cheaper, Kimi K2 (Jul 2025) 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 (Jul 2025) lists at $0.57 and $2.30.

### Is Kimi K2 (Jul 2025) or Llama-3.3-70B-Instruct better for coding?

Kimi K2 (Jul 2025) scores higher on coding benchmarks: 43.1 versus 30.6 in the Noometry coding category.

### Which has the bigger context window?

Kimi K2 (Jul 2025) does, with 262K tokens against 128K.

### How many benchmarks do Kimi K2 (Jul 2025) and Llama-3.3-70B-Instruct share?

25 benchmarks have published results for both models. Kimi K2 (Jul 2025) has 42 scored results on Noometry and Llama-3.3-70B-Instruct has 43.