Model Profile

Kimi K2 Thinking

Name: Kimi K2 Thinking
Author: kimi

External Benchmark Shadowexternal_benchmark_shadowpublic

4,096 ctx

Use this page to decide where this model is a strong fit. Rankings below are benchmark-backed by use case, with explicit confidence and contributor metrics.

Identity

ID: external/kimi/kimi-k2-thinking

Author: kimi

Origin: external_benchmark_shadow

Arch: unknown

Benchmark Coverage

Scored use cases: 12

Avg confidence: 31.3%

Evidence points: 209

Raw rows: 487

Weighted rows: 34

Catalog Metadata

Parameters: unknown

Context window: 4096

Downloads: 0

Coverage Diagnostics

actively scored

Use-Case Scores

113

Total Measurements

487

Weighted Measurements

Weighted Sources

Raw Source Coverage

vals_mmlu_pro 60vals_mgsm 48vals_finance_agent 40corpfin_taxeval_public 28vals_medqa 28vals_legal_bench 24

Weighted Source Coverage

openhands_index 5vals_finance_agent 5sonar_java_quality 4halluhard_leaderboard 3vals_corp_fin_v2 3openhands_issue_resolution 2

Best Use Cases for This Model

Use Case	Vertical	Score	Confidence	Evidence	Top Contributor
Quant research code generation use_case.fin.alpha_research_codegen	finance	22.6%	44.1%	30	Sonar Java Quality Leaderboard: functional_skill_pct
Verilog/VHDL generation use_case.eda.verilog_generation	engineering	21.6%	31.0%	15	Sonar Java Quality Leaderboard: functional_skill_pct
Refactoring assistant use_case.dev.refactoring	developer_tools	21.3%	30.7%	15	Sonar Java Quality Leaderboard: functional_skill_pct
Thesis red teaming use_case.fin.thesis_red_team	finance	21.3%	35.6%	18	Vals CorpFin v2: overall_accuracy_pct
Debugging assistant use_case.dev.debugging	developer_tools	20.1%	29.1%	15	Sonar Java Quality Leaderboard: functional_skill_pct
Earnings call synthesis use_case.fin.earnings_call_synthesis	finance	19.1%	32.0%	18	Vals CorpFin v2: overall_accuracy_pct
Accounts payable invoice extraction (text) use_case.fin.ap_invoice_extraction	finance	19.1%	30.2%	18	Vals CorpFin v2: overall_accuracy_pct
Transaction anomaly narrative use_case.fin.transaction_anomaly_narrative	finance	18.7%	31.3%	18	Vals CorpFin v2: overall_accuracy_pct
KYC profile synthesis use_case.fin.kyc_profile_synthesis	finance	17.9%	30.0%	18	Vals CorpFin v2: overall_accuracy_pct
AML alert triage use_case.fin.aml_alert_triage	finance	17.9%	30.0%	18	Vals CorpFin v2: overall_accuracy_pct
Integration test generation use_case.dev.integration_tests	developer_tools	17.5%	25.2%	14	Sonar Java Quality Leaderboard: functional_skill_pct
Simulation setup assistant use_case.eng.simulation_setup_assistant	engineering	16.8%	26.0%	12	Vals SWE-bench: overall_accuracy_pct

Deployment Fit Calculator

Model

Kimi K2 Thinking

external/kimi/kimi-k2-thinking

Hardware

Quantization: 4-bit

2-bit8-bit

Insufficient

Unknown parameter count. Cannot estimate deployment fit.

Required VRAM

~0.0GB

Est. Throughput

0.00 tok/s

Deployment Fit Matrix

GPU	4-bit	6-bit	8-bit
RTX 3060 12GB	Insufficient	Insufficient	Insufficient
RTX 3090 24GB	Insufficient	Insufficient	Insufficient
RTX 4090 24GB	Insufficient	Insufficient	Insufficient
Mac Studio M2 Ultra 192GB	Insufficient	Insufficient	Insufficient