Kimi K3: score and benchmark evidence
Historical release · · Moonshot AI
104.9 points · Rank 8 among 19 ranked models · Stability range 95.4–113.7 points.
Rank follows the point estimate; it does not establish statistically significant superiority. The scale is anchored at mean 100, SD 15 in the frozen calibration cohort, not human IQ or a percentage.
One score summarizes demonstrated capability across published reasoning settings and seven equally weighted domains. Related benchmark variants share a family budget, and influence from one evaluator is capped.
Release tli-2026-v1.0-2026-09-10-benchmarks-1. Methodology and limitations for this release.
This page preserves historical evidence. See the current model record.
Capability domains
Domain scores use calibration-cohort standard-deviation units. Unmeasured domains are shown as unavailable, not zero.
| Domain | Score (SD units) | Observed benchmarks |
|---|---|---|
| Knowledge & reasoning | -0.10 | 14 |
| Coding & software engineering | 0.15 | 11 |
| Agentic tool & computer use | -0.02 | 11 |
| Professional & real-world work | 0.03 | 16 |
| Multimodal & vision | 0.20 | 8 |
| Cybersecurity | Unavailable | 0 |
| Preference & communication | 0.52 | 4 |
Published benchmark evidence
64 results contribute to this score, from 93 collected results and 60 contributing benchmark families. Counts are not independent sample sizes.
Results, source citations, review decisions and compatible reasoning modes are stored in a versioned relational database. Each publication uses a sealed, checksum-verified export; CSV files are optional exports, not editable scoring inputs. Previous revisions remain available for audit. Collected means a numeric result exists in the database. Used means its source, exact model checkpoint and benchmark protocol have been reviewed and its benchmark has a frozen calibration supported by multiple providers. For each benchmark we select the highest published aggregate across comparable reasoning modes, not necessarily maximum effort, and count it once. More tested settings provide more chances to record a high score; this selection advantage is not corrected by the conditional interval. Developer-reported and sole/unspecified settings remain eligible and disclosed. Unresolved does not mean false or useless: conflicting metrics, unverified citations and unmatched versions stay collected until reconciled.
The score stability range is a 95% bootstrap score interval conditional on frozen calibration and comparable evidence availability. It resamples benchmark families and evaluators and adds a frozen missing-evidence correction. It excludes selection across reasoning settings, selective publication, calibration changes and new domains. It is not a 95% guarantee about true capability or where a future score will land. Sparse or conflicting evidence can widen the range.
Reporting sensitivity: ±1.1 points. Sensitivity to selectively published results. This stress range is not a confidence interval or a probability statement.
GDPval-AA v2 (Elo): 1587.5 native-points
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (max); Artificial Analysis published evaluation harness
Metric: Elo
Original source · Retrieved 2026-09-06
Best published result among 2 reasoning configurations of the exact AA model release; fallback configurations excluded.
Terminal-Bench 2.1: 85.018727%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (max); Artificial Analysis published evaluation harness
Metric: pass@1, AA harness
Original source · Retrieved 2026-09-06
Best published result among 2 reasoning configurations of the exact AA model release; fallback configurations excluded.
DeepSWE v1.1: 68.514412%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: kimi-k3; max; mini-swe-agent
Metric: pass@1 over scored rollout attempts
Original source · Retrieved 2026-09-06
Selected the best pass@1 across published reasoning efforts within mini-swe-agent. Pass@4 is not substituted.
LMArena (Chatbot Arena) — Text: 1488.665886 native-points
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: kimi-k3-max; Arena text overall; style control enabled
Metric: style-controlled text overall Elo
Original source · Retrieved 2026-09-06
Best published reasoning-mode rating within the same text overall board. Different leaderboard slices and undated earlier DeepSeek checkpoints are not pooled.
AA Intelligence Index v4.1.1 (score): 60 native-points
Not used in the score. Source category not recorded. Review status: excluded.
Held-out composite, never a scoring input
GPQA Diamond: 93.535354%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (max); Artificial Analysis published evaluation harness
Metric: accuracy, AA
Original source · Retrieved 2026-09-06
Best published result among 2 reasoning configurations of the exact AA model release; fallback configurations excluded.
AA-LCR: 74.7%
Not used in the score. Source category: independent. Review status: unresolved.
Configuration: Kimi K3 (max)
Original source · Retrieved 2026-09-06
Current AA result is LCR v1.1. The inherited AA-LCR row has no reviewed version match; a new version cannot silently substantiate the older value.
Finance Agent v2 — Vals: 54.4%
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Vals AI Finance Agent v2 harness
Metric: Task success
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
Excel Modeling Benchmark — Vals: 66.4%
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Vals AI Excel Modeling overall evaluation
Metric: Accuracy
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
SciCode: 59.490741%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (max); Artificial Analysis published evaluation harness
Metric: accuracy, AA harness
Original source · Retrieved 2026-09-06
Best published result among 2 reasoning configurations of the exact AA model release; fallback configurations excluded.
Humanity's Last Exam (no tools): 46.895273%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (max); Artificial Analysis published evaluation harness
Metric: accuracy, no tools, AA
Original source · Retrieved 2026-09-06
Best published result among 2 reasoning configurations of the exact AA model release; fallback configurations excluded.
Legal Research Bench — Vals: 44.23%
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Vals AI Legal Research overall harness
Metric: Task success
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
Harvey LAB (Vals): 10.83%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 10.83 for Kimi K3 on Harvey LAB (Vals) still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
AA-Briefcase Overall Elo: 1499.88 native-points
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (max); Artificial Analysis published evaluation harness
Metric: overall Elo
Original source · Retrieved 2026-09-06
Best published result among 1 reasoning configurations of the exact AA model release; fallback configurations excluded.
MedScribe — Vals: 87.96%
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Vals AI health evaluation
Metric: Accuracy
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
SWE-bench Verified — Vals: 93.4%
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Vals AI mini-SWE-agent bash harness
Metric: Resolved issue rate
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
LegalBench — Vals: 86.02%
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Vals AI LegalBench evaluation
Metric: Accuracy
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
TaxEval v2 — Vals: 75.72%
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Vals AI max-compute evaluation
Metric: Accuracy
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
KingBench 3: 77.5%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 77.5 for Kimi K3 on KingBench 3 still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
Vals Index: 74.7%
Not used in the score. Source category: independent. Review status: excluded.
Configuration: kimi-k3; max effort; Vals AI index methodology
Metric: Composite score
Original source · Retrieved 2026-08-16T00:00:00.000Z
Held-out composite, never a scoring input
Public Benefits Bench — Vals: 68.27%
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Vals AI public-benefits evaluation
Metric: Accuracy
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
MedCode — Vals: 48.88%
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Vals AI health evaluation
Metric: Accuracy
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
AutomationBench v1.0.6: 46.7%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 46.7 for Kimi K3 on AutomationBench v1.0.6 still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
AA-Omniscience: 19.7 native-points
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (max); Artificial Analysis published evaluation harness
Metric: knowledge reliability index -100 to 100
Original source · Retrieved 2026-09-06
Best published result among 2 reasoning configurations of the exact AA model release; fallback configurations excluded.
MMLU Pro — Vals: 87.97%
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Vals AI five-shot MMLU-Pro harness
Metric: Accuracy
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
LiveBench: 79.193048%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: kimi-k3; LiveBench 2026-06-25 release
Metric: official overall: equal mean of seven category means
Original source · Retrieved 2026-09-06
Task mapping archived from categories_2026_06_25.json. Aggregate follows the official UI, not a flat mean of the 23 tasks. Undated old DeepSeek checkpoints excluded.
CorpFin v2 — Vals: 71.56%
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Vals AI CorpFin v2 evaluation with Sonnet 4.5 judge
Metric: Accuracy
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
MortgageTax — Vals: 66.34%
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Vals AI max-compute evaluation
Metric: Accuracy
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
SAGE — Vals: 54.26%
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Vals AI max-compute evaluation
Metric: Score
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
Code Migration †: 16.1%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 16.1 for Kimi K3 on Code Migration † still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
Toolathlon-Verified: 76.5%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi Kimi K3 (max) ✓; Default agent
Metric: pass@1, verified tasks
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
Agents' Last Exam: 28.3%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3; Max; Kimi Code
Metric: pass rate, ALE-V1 147 tasks; not partial score
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
Terminal-Bench v3.0: 17.4%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 17.4 for Kimi K3 on Terminal-Bench v3.0 still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
LiveCodeBench: 87.19%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 87.19 for Kimi K3 on LiveCodeBench still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
CyberGym: 80%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 80 for Kimi K3 on CyberGym still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
Humanity's Last Exam w/ tools: 59.8%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 59.8 for Kimi K3 on Humanity's Last Exam w/ tools still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
Design Arena (Elo): 1453 native-points
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Design Arena 2026-08-11 rating snapshot
Metric: Elo rating
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
Vals Multimodal Index: 73.42%
Not used in the score. Source category: independent. Review status: excluded.
Configuration: kimi-k3; max effort; Vals AI multimodal index methodology
Metric: Composite score
Original source · Retrieved 2026-08-16T00:00:00.000Z
Held-out composite, never a scoring input
SWE-Marathon v1.1: 48.1%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 48.1 for Kimi K3 on SWE-Marathon v1.1 still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
ProgramBench v1 (Raw Pass Rate) — Vals: 62.768%
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi/kimi-k3 · Vals published row (effort not stated)
Metric: Raw Pass Rate: mean percent of hidden behavioral tests passed per task. Not Fully Resolved (all tests pass) and not Almost Resolved (≥95% of tests pass).
Original source · Retrieved 2026-09-05
Matched exact metric/value and the archived benchmark-specific configuration review. Unknown native effort stays explicitly unknown.
FrontierSWE: 81.2%
Contributes to the score. Developer-reported. Review status: reviewed-developer.
Configuration: Kimi K3; max; Kimi developer comparison table; benchmark-specific configuration
Metric: FrontierSWE
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
FrontierMath v2 Tier 4: 39.02439%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: kimi-k3_max; task version 2.0.0
Metric: mean score percentage, FrontierMath-Tier-4-v2-Private
Original source · Retrieved 2026-09-06
Exact Epoch task version; selected best mean score across reported reasoning configurations. No pooling of task versions or substitution of best scorer output.
CursorBench v3.2: 60.8%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 Max; Cursor harness
Metric: CursorBench 3.2 score percentage
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
APEX-Agents Mean Criteria Passed: 55.4%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: kimi-k3; max; loop_truncated_tools_agent
Metric: mean-score
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
CharXiv (with CI / RQ): 91.3%
Contributes to the score. Developer-reported. Review status: reviewed-developer.
Configuration: Kimi K3; max; Kimi developer comparison table; with Python tools
Metric: CharXiv (RQ)
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
APEX-SWE (Pass@1, Terminus-2): 48%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: kimi-k3; max; terminus-2
Metric: pass-1
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
BrowseComp: 91.2%
Contributes to the score. Developer-reported. Review status: reviewed-developer.
Configuration: Kimi K3; max; Kimi developer comparison table; benchmark-specific configuration
Metric: BrowseComp
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
MMMU-Pro: 80.520231%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (max); Artificial Analysis published evaluation harness
Metric: accuracy, AA
Original source · Retrieved 2026-09-06
Best published result among 2 reasoning configurations of the exact AA model release; fallback configurations excluded.
ARC-AGI-2: 60.4%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3; Max; ARC Prize standard harness
Metric: semi-private verified score percentage
Original source · Retrieved 2026-09-06
Best reasoning variant within the verified standard-harness summary. Provider Adapter, public tasks and community agent scores are not substituted.
OSWorld 2.0 — Anthropic H2H: 58.3%
Contributes to the score. Source category: third_party. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; Anthropic Opus 5 published head-to-head setup
Metric: Task success
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
ExploitBench: 32.2%
Not used in the score. Source category not recorded. Review status: excluded.
Inherited comparator metric/scaffold equivalence to Astra Cap Percent is not established; the source flags possible contamination. Excluded pending protocol audit.
PostTrainBench: 36.6%
Contributes to the score. Developer-reported. Review status: reviewed-developer.
Configuration: Kimi K3; max; Kimi developer comparison table; benchmark-specific configuration
Metric: PostTrainBench
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
FrontierCode v1.1 Extended: 58.19%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3; none; mini-swe-agent
Metric: v1.1 Extended cost-adjusted new_score, 150 tasks
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
JobBench: 54.3%
Contributes to the score. Developer-reported. Review status: reviewed-developer.
Configuration: Kimi K3; max; Kimi developer comparison table; benchmark-specific configuration
Metric: JobBench
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
BabyVision (with CI): 85.7%
Contributes to the score. Developer-reported. Review status: reviewed-developer.
Configuration: Kimi K3; max; Kimi developer comparison table; with Python tools
Metric: BabyVision w/ python
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
PerceptionBench: 58.5%
Contributes to the score. Developer-reported. Review status: reviewed-developer.
Configuration: Kimi K3; max; Kimi developer comparison table; benchmark-specific configuration
Metric: PerceptionBench
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
OSWorld-Verified: 84.8%
Contributes to the score. Developer-reported. Review status: reviewed-developer.
Configuration: Kimi K3; max; Kimi developer comparison table; benchmark-specific configuration
Metric: OSWorld-Verified
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
Creative Writing v3: 2070.8 native-points
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: kimi-k3; published evaluation, effort unspecified
Metric: Creative Writing v3 Elo
Original source · Retrieved 2026-09-06
Exact checkpoint on the evaluator-hosted board. A starred model is externally submitted; it is not claimed as independently rerun by the benchmark owner.
MathArena — ArXivMath Jun 2026: 72.11%
Contributes to the score. Source category: benchmark_org. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; MathArena June 2026 competition harness
Metric: Accuracy
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
MathArena — BrokenArXiv Jun 2026: 51.85%
Contributes to the score. Source category: benchmark_org. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; MathArena June 2026 competition harness
Metric: Accuracy
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
AIIQ Composite IQ: 122 native-points
Not used in the score. Source category: independent. Review status: excluded.
Configuration: kimi-k3; max effort; AIIQ composite snapshot
Metric: Composite IQ
Original source · Retrieved 2026-08-16T00:00:00.000Z
Held-out composite, never a scoring input
MCP Atlas †: 84.2%
Contributes to the score. Developer-reported. Review status: reviewed-developer.
Configuration: Kimi K3; max; Kimi developer comparison table; benchmark-specific configuration
Metric: MCP-Atlas
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
OfficeQA Pro †: 63.3%
Contributes to the score. Developer-reported. Review status: reviewed-developer.
Configuration: Kimi K3; max; Kimi developer comparison table; benchmark-specific configuration
Metric: OfficeQA Pro
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
EQ-Bench 4: 1339.3 native-points
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: moonshotai/kimi-k3; evaluator published configuration, effort unspecified
Metric: EQ-Bench 4 Elo
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
NL2Repo: 58%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 58 for Kimi K3 on NL2Repo still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
τ²-Bench Telecom: 80.63%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 80.63 for Kimi K3 on τ²-Bench Telecom still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
FrontierCode v1.1 Main — Anthropic H2H: 44.2%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 44.2 for Kimi K3 on FrontierCode v1.1 Main — Anthropic H2H still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
MathArena — ArXivMath May 2026: 61.67%
Contributes to the score. Source category: benchmark_org. Review status: reviewed-archive.
Configuration: kimi-k3; max effort; MathArena May 2026 competition harness
Metric: Accuracy
Original source · Retrieved 2026-08-16T00:00:00.000Z
Exact-value match to an existing verified, comparable, pinned-configuration observation. Archived review retained; not asserted newly independently replicated.
τ³-Banking: 45.979381%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (max); Artificial Analysis published evaluation harness
Metric: strict task success, AA
Original source · Retrieved 2026-09-06
Best published result among 2 reasoning configurations of the exact AA model release; fallback configurations excluded.
CritPt: 23.428571%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (max); Artificial Analysis published evaluation harness
Metric: accuracy, AA official grading
Original source · Retrieved 2026-09-06
Best published result among 2 reasoning configurations of the exact AA model release; fallback configurations excluded.
GDP.pdf: 19%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (Max reasoning); Surge original evaluation
Metric: all-pass percentage, original Surge GDP.pdf
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
ProofBench v1.1 — Vals: 87%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 87 for Kimi K3 on ProofBench v1.1 — Vals still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
Vibe Code Bench v1.1 — Vals: 84.96%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 84.96 for Kimi K3 on Vibe Code Bench v1.1 — Vals still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
MineBench: 1812 native-points
Not used in the score. Source category not recorded. Review status: excluded.
Unresolved Pro/base model identities in source catalog; exclude until matched.
ARC-AGI-1: 94.5%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3; Max; ARC Prize standard harness
Metric: semi-private verified score percentage
Original source · Retrieved 2026-09-06
Best reasoning variant within the verified standard-harness summary. Provider Adapter, public tasks and community agent scores are not substituted.
Harvey LAB-AA: 94.64%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 94.64 for Kimi K3 on Harvey LAB-AA still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
EnterpriseOps-Gym-AA: 45.33%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 45.33 for Kimi K3 on EnterpriseOps-Gym-AA still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
MLCR-AA (Medical Long Context Reasoning): 38.333333%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (max); Artificial Analysis published evaluation harness
Metric: overall accuracy, AA
Original source · Retrieved 2026-09-06
Best published result among 1 reasoning configurations of the exact AA model release; fallback configurations excluded.
AA-AnalystAgent: 38.75%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (max); Artificial Analysis published evaluation harness
Metric: accuracy, AA
Original source · Retrieved 2026-09-06
Best published result among 1 reasoning configurations of the exact AA model release; fallback configurations excluded.
ITBench-AA: 47.693032%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (max); Artificial Analysis published evaluation harness
Metric: root cause accuracy, AA
Original source · Retrieved 2026-09-06
Best published result among 1 reasoning configurations of the exact AA model release; fallback configurations excluded.
DeepSearchQA: 95%
Not used in the score. Developer-reported. Review status: reviewed-developer.
Configuration: Kimi K3; max; Kimi developer comparison table; benchmark-specific configuration
Metric: DeepSearchQA (F1)
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
Vending-Bench 2: 5165.035 native-points
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: Kimi K3 (Moonshot); Andon Labs Vending-Bench 2
Metric: arithmetic mean final bank balance USD after one simulated year
Original source · Retrieved 2026-09-06
Compare reasoning configurations using their mean over runs. Geometric mean, individual best runs, Fireworks Kimi endpoint, Vending Arena, and April DeepSeek Pro are not substituted.
IFBench: 74.4%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 74.4 for Kimi K3 on IFBench still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
CWE-bench: 22.5%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 22.5 for Kimi K3 on CWE-bench still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
ZeroBench: 23%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 23 for Kimi K3 on ZeroBench still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
MathVision (with CI): 97.8%
Contributes to the score. Developer-reported. Review status: reviewed-developer.
Configuration: Kimi K3; max; Kimi developer comparison table; with Python tools
Metric: MathVision
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
OmniDocBench 1.5: 91.1%
Contributes to the score. Developer-reported. Review status: reviewed-developer.
Configuration: Kimi K3; max; Kimi developer comparison table; benchmark-specific configuration
Metric: OmniDocBench
Original source · Retrieved 2026-09-06
Reconciled against the exact model/configuration in the evaluator’s public source. Updated source value recorded explicitly; best published reasoning mode selected within the same benchmark protocol.
Mystery Game Puzzles v1.0.4 — Epoch: 26%
Contributes to the score. Source category: independent. Review status: reviewed-independent.
Configuration: kimi-k3_max; task version 1.0.4
Metric: mean score percentage, Mystery Game Puzzles
Original source · Retrieved 2026-09-06
Exact Epoch task version; selected best mean score across reported reasoning configurations. No pooling of task versions or substitution of best scorer output.
BioMysteryBench v1 — Vals: 71.481%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 71.481 for Kimi K3 on BioMysteryBench v1 — Vals still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
Tax Agent Bench v1 — Vals: 68.672%
Not used in the score. Source category not recorded. Review status: unresolved.
Not admitted after the September 6 model review: the collected 68.672 for Kimi K3 on Tax Agent Bench v1 — Vals still lacks a matched source entry establishing its checkpoint and benchmark-specific metric, tools and retry setup. This is a citation/comparability gap, not a finding that the result is false or useless. More or maximum reasoning effort is not required.
RuneBench (30m, mean ln(1 + XP/min)): 3.753577 native-points
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi3
Metric: Arithmetic mean of ln(1 + peakXpRate) over all 16 skills, exactly as the owner’s Heatmap.js default ranking. Data rates are already normalized; do not divide by 200 again.
Original source · Retrieved 2026-09-06
Matched exact metric/value and the archived benchmark-specific configuration review. Unknown native effort stays explicitly unknown.
VoxelBench (text-to-voxel, Glicko-2): 1932 native-points
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: Kimi K3 (Max)
Metric: Glicko-2 rating from the text-to-voxel board; not win percentage, deviation, image-input mode or MineBench Elo.
Original source · Retrieved 2026-09-06
Matched exact metric/value and the archived benchmark-specific configuration review. Unknown native effort stays explicitly unknown.
Code Arena WebDev (overall): 1674 native-points
Contributes to the score. Source category: independent. Review status: reviewed-archive.
Configuration: kimi-k3-max
Metric: Published overall WebDev preference rating, not task success percentage or Arena Text score.
Original source · Retrieved 2026-09-06
Matched exact metric/value and the archived benchmark-specific configuration review. Unknown native effort stays explicitly unknown.
Cite this model record
The Latent. “Kimi K3: score and benchmark evidence.” 2026-09-10. Release tli-2026-v1.0-2026-09-10-benchmarks-1.
Download release data and definitions (JSON) · Data usage terms
Related reporting
These dated stories provide reporting context; they may describe an earlier release.