Data as of October 29, 2026 · How the score is built
Released Oct 23, 20261.05M contextModel card
Kimi K3.1 Pro
Decision readingKimi K3.1 Pro scores 74.6 out of 100 and ranks #13 of 216. This profile shows 47 source-displayable benchmark rows; its strongest eligible category is Multimodal & Grounded at #1. API pricing is $2 input and $10 output per million tokens, with cached input at $0.2.
Released Oct 23, 2026 — see all recent releases
Decision snapshot
Each value carries a field reference instead of floating alone. Markers compare this model with the current ranked and priced catalog; they are not absolute quality thresholds.
Capability
74.6/100
field median 50.5#13 of 216 ranked models
Public
#13of 216
Verified —
Price
$2input / $10 output
input median $0.75cached $0.20 · blended $6
Speed
43tok/s
field median 92 tok/sFirst token 49.6 s
Context
1.05Mtokens
field median 256,000Reported for this model; direct source link not stored
Strongest published evidence
Multimodal & Grounded ranks #1. Particularly strong for screenshots, documents, charts, and grounded multimodal workflows.
Validate before choosing
47 published rows leave some tracked benchmark slots empty. Knowledge is its lowest eligible category at #13.
Source-linked · 47 displayable benchmark rows
Category score record
Scores and ranks appear only where published evidence can be displayed. The table keeps the score, weight, cohort, and evidence state together.
| Category | Score | Rank | Percentile | Weight | Benchmarks | Evidence |
|---|---|---|---|---|---|---|
| AgenticRank #9 of 135Percentile 94thWeight 22%18 benchmarksVerified | 73.2 | 18 benchmarks | Verified | |||
| CodingWeight 20%7 benchmarksVerified | Score pending | 7 benchmarks | Verified | |||
| ReasoningWeight 17%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| MultimodalRank #1 of 58Percentile 100thWeight 12%15 benchmarksVerified | 100.0 | 15 benchmarks | Verified | |||
| KnowledgeRank #13 of 179Percentile 93rdWeight 12%5 benchmarksVerified | 75.1 | 5 benchmarks | Verified | |||
| MultilingualWeight 7%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| Inst. FollowingWeight 5%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| MathWeight 5%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured |
47 of 782 tracked benchmark slots have displayable evidence · bars run 0–100
Coverage detailsHow much of this is verified
Coverage is split by category so a strong number never hides a thin evidence base. Verified means the row is tied to a published source; provisional rows remain visible but separate.
- Agentic18/18 verified
- Coding7/7 verified
- ReasoningNot measured
- Multimodal15/15 verified
- Knowledge5/5 verified
- MultilingualNot measured
- Inst. FollowingNot measured
- MathNot measured
Capability shape
Each axis shows percentile within that category’s eligible cohort. The comparison outline is the median of the six nearest public-score peers; a collapsed vertex means the category is not rank-eligible.
Kimi K3.1 Pro category percentile values
- Agentic94th percentile
- CodingNot eligible
- ReasoningNot eligible
- Multimodal100th percentile
- Knowledge93rd percentile
- MultilingualNot eligible
- Instruction followingNot eligible
- MathNot eligible
The dashed outline is median of 6 nearest peers.
Eligible category ranks
- Agentic#9/135
- CodingNot ranked
- ReasoningNot ranked
- Multimodal#1/58
- Knowledge#13/179
- MultilingualNot ranked
- Inst. FollowingNot ranked
- MathNot ranked
Benchmark ledger
Coding opens by default. The marker compares each value with the best source-verified result in the catalog; provisional leaders do not set the reference. Expand the remaining categories for every published row.
Coding7 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| DeepSWE | Score75.5% | Versus best verified row Best verified: Muse Wonder Full · 79.0% | Gap3.5 behind | WeightDisplay only | Provider exact |
| ProgramBench (binary)ProgramBench binary score (task counts only at 95% or higher) | Score31.5% | Versus best verified row Best verified: Kimi K3.1 Pro · 31.5% | GapBest verified | WeightDisplay only | Provider exact |
| Kimi Code Bench v2 | Score77.1% | Versus best verified row Best verified: Kimi K3.1 Pro · 77.1% | GapBest verified | WeightDisplay only | Provider exact |
| Kimi Code Bench 3.0 Long-horizonKimi Code Bench 3.0 Long-horizon (Internal) | Score61.2% | Versus best verified row Best verified: Kimi K3.1 Pro · 61.2% | GapBest verified | WeightDisplay only | Provider exact |
| sweMarathon | Score51.7% | Versus best verified row Best verified: Step 5 Preview · 72.7% | Gap21 behind | WeightDisplay only | Provider exact |
| PostTrainBench v1.2 | Score36.5% | Versus best verified row Best verified: Kimi K3.1 Pro · 36.5% | GapBest verified | WeightDisplay only | Provider exact |
| Bug Hunt Bench | Score36.0 fixes | Versus best verified row Best verified: Claude Sonnet 5.5 · 51.3 fixes | Gap15.3 behind | WeightDisplay only | Benchmark exact |
Agentic18 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| BrowseComp | Score92.1% | Versus best verified row Best verified: Atria Dawn Preview · 92.5% | Gap0.4 behind | Weight8% ref. weight | Provider exact |
| Terminal-Bench 4.0 | Score65.70% | Versus best verified row Best verified: Claude Sonnet 5.5 · 70.60% | Gap4.9 behind | WeightDisplay only | Provider exact |
| HLE w/ toolsHumanity's Last Exam with tools | Score70.0% | Versus best verified row Best verified: Kimi K3.1 Pro · 70.0% | GapBest verified | WeightDisplay only | Provider exact |
| Agents' Last Exam | Score43.2% | Versus best verified row Best verified: GPT-6.1 Terra · 60.0% | Gap16.8 behind | WeightDisplay only | Provider exact |
| Terminal-Bench-Science 0.1 | Score38.1% | Versus best verified row Best verified: GPT-6 Astra · 64.6% | Gap26.5 behind | WeightDisplay only | Provider exact |
| BioMysteryBenchBioMysteryBench (overall, v11 release) | Score80.6% | Versus best verified row Best verified: Kimi K3.1 Pro · 80.6% | GapBest verified | WeightDisplay only | Provider exact |
| Agents' Last Exam (CLI)Agents' Last Exam v1.1, CLI subset | Score44.2% | Versus best verified row Best verified: Kimi K3.1 Pro · 44.2% | GapBest verified | WeightDisplay only | Provider exact |
| APEX-Agents | Score67.2% | Versus best verified row Best verified: Kimi K3.1 Pro · 67.2% | GapBest verified | WeightDisplay only | Provider exact |
| Frontier Finance | Score58.6% | Versus best verified row Best verified: Kimi K3.1 Pro · 58.6% | GapBest verified | WeightDisplay only | Provider exact |
| HANDBOOK.md | Score44.6% | Versus best verified row Best verified: Kimi K3.1 Pro · 44.6% | GapBest verified | WeightDisplay only | Provider exact |
| OSWorld 2.0 | Score79.6% | Versus best verified row Best verified: Kimi K3.1 Pro · 79.6% | GapBest verified | WeightDisplay only | Provider exact |
| Business OperationBusiness Operation (Internal) | Score37.8% | Versus best verified row Best verified: Kimi K3.1 Pro · 37.8% | GapBest verified | WeightDisplay only | Provider exact |
| Financial ModelingFinancial Modeling (Internal) | Score57.6% | Versus best verified row Best verified: Kimi K3.1 Pro · 57.6% | GapBest verified | WeightDisplay only | Provider exact |
| Financial ResearchFinancial Research (Internal) | Score70.4% | Versus best verified row Best verified: Kimi K3.1 Pro · 70.4% | GapBest verified | WeightDisplay only | Provider exact |
| Kimi Hallu Bench Hard (tools)Kimi Hallucination Benchmark Hard with tools (Internal) | Score8.3% | Versus best verified row Best verified: Kimi K3.1 Pro · 8.3% | GapBest verified | WeightDisplay only | Provider exact |
| ExploitGym | Score20.7% | Versus best verified row Best verified: Gemini 4 Carbon · 53.5% | Gap32.8 behind | WeightDisplay only | Provider exact |
| MazeBenchMazeBench: visual spatial reasoning in a 3D open world | Score2% | Versus best verified row Best verified: GPT-6 Astra · 23% | Gap21 behind | WeightDisplay only | Benchmark exact |
| OSWorld 2.1OSWorld 2.1 (all 108 tasks) | Score79.6% | Versus best verified row Best verified: Claude Opus 5.5 · 81.8% | Gap2.2 behind | WeightDisplay only | Provider exact |
Multimodal15 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| MMMU-ProMassive Multi-discipline Multimodal Understanding Pro | Score84.1% | Versus best verified row Best verified: GPT-6.1 Terra · 85.5% | Gap1.4 behind | WeightWeighted 40% | Provider exact |
| OfficeQA Pro | Score67.4% | Versus best verified row Best verified: Claude Opus 5.5 · 67.7% | Gap0.3 behind | WeightWeighted 25% | Provider exact |
| CharXivCharXiv Reasoning | Score93.4% | Versus best verified row Best verified: Qwen3.8 Max · 93.5% | Gap0.1 behind | WeightWeighted 20% | Provider exact |
| OfficeQA Pro v2 | Score64.4% | Versus best verified row Best verified: Kimi K3.1 Pro · 64.4% | GapBest verified | WeightDisplay only | Provider exact |
| BenchCAD Vision2Code (no tools)BenchCAD Vision2Code voxel IoU without tools | Score0.728 | Versus best verified row Best verified: Claude Sonnet 5.5 · 0.747 | Gap0 behind | WeightDisplay only | Provider exact |
| BenchCAD Vision2Code (tools)BenchCAD Vision2Code voxel IoU with tools | Score0.927 | Versus best verified row Best verified: Claude Sonnet 5.5 · 0.963 | Gap0 behind | WeightDisplay only | Provider exact |
| MMMU-Pro w/ PythonMMMU-Pro with Python | Score90% | Versus best verified row Best verified: Kimi K3.1 Pro · 90% | GapBest verified | WeightDisplay only | Provider exact |
| CharXiv w/o toolsCharXiv Reasoning without tools | Score89% | Versus best verified row Best verified: Kimi K3.1 Pro · 89% | GapBest verified | WeightDisplay only | Provider exact |
| Chartography (tools)Chartography with image and code tools | Score90.2% | Versus best verified row Best verified: Claude Sonnet 5.5 · 90.2% | GapBest verified | WeightDisplay only | Provider exact |
| BabyVision w/ PythonBabyVision with Python | Score95.4% | Versus best verified row Best verified: Kimi K3.1 Pro · 95.4% | GapBest verified | WeightDisplay only | Provider exact |
| ZeroBench | Score39.0% | Versus best verified row Best verified: Muse Spark Open · 50.0% | Gap11 behind | WeightDisplay only | Provider exact |
| ZeroBench w/ PythonZeroBench_main with Python | Score63.0% | Versus best verified row Best verified: Kimi K3.1 Pro · 63.0% | GapBest verified | WeightDisplay only | Provider exact |
| GDP.pdf (Gemini judge)GDP.pdf strict pass rate, page images plus text, Gemini judge | Score34.6% | Versus best verified row Best verified: Kimi K3.1 Pro · 34.6% | GapBest verified | WeightDisplay only | Provider exact |
| GDP.pdf w/ Python (Gemini judge)GDP.pdf task-macro rubric pass rate with Python, Gemini judge | Score85.7% | Versus best verified row Best verified: Kimi K3.1 Pro · 85.7% | GapBest verified | WeightDisplay only | Provider exact |
| PerceptionBenchPerceptionBench (Internal) | Score68.3% | Versus best verified row Best verified: Kimi K3.1 Pro · 68.3% | GapBest verified | WeightDisplay only | Provider exact |
Knowledge5 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| HLEHumanity's Last Exam | Score70% | Versus best verified row Best verified: Kimi K3.1 Pro · 70% | GapBest verified | Weight44% ref. weight | Provider exact |
| GPQAGraduate-Level Google-Proof Q&A | Score94.2% | Versus best verified row Best verified: GPT-6 Astra · 96% | Gap1.8 behind | Weight3% ref. weight | Provider exact |
| HLE w/o toolsHumanity's Last Exam without tools | Score51.1% | Versus best verified row Best verified: Claude Opus 5.5 · 64.4% | Gap13.3 behind | WeightDisplay only | Provider exact |
| GPQA-DGPQA Diamond | Score94.2% | Versus best verified row Best verified: GPT-6 Astra · 96.0% | Gap1.8 behind | WeightDisplay only | Provider exact |
| Kimi Hallu BenchKimi Hallucination Benchmark without tools (Internal) | Score5.4% | Versus best verified row Best verified: Kimi K3.1 Pro · 5.4% | GapBest verified | WeightDisplay only | Provider exact |
External signals2 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| ExploitBenchExploitBench v8-bench | Score75% | Versus best verified row Best verified: GPT-6 Astra · 100% | Gap25.3 behind | WeightDisplay only | Provider exact |
| xdev-bench-hardxdev-bench-hard (Internal, pass@3) | Score81.0% | Versus best verified row Best verified: Kimi K3.1 Pro · 81.0% | GapBest verified | WeightDisplay only | Provider exact |
What it costs to get this score
Published API price against the public score. The x-axis uses a log scale; the dashed path marks models that are not beaten by a cheaper, higher-scoring option. Price uses average of published input and output rates.
Kimi K3.1 Pro · 74.6 score · $6 blended per million tokens
The chart opens on the current model. Scroll horizontally to inspect the full price axis.
Horizontal: blended price per million tokens, log scale · Vertical: public score
Lineage
The sequence follows explicit supersedes links. Each score is estimated for that model; a relative can inform a sparse estimate but never sets a floor, so a newer release can score below an earlier one. Scores and prices remain blank when the corresponding public row or first-party rate is unavailable.
Base entry
Spec sheet
Each documented value carries its source. Missing fields stay visible as not sourced or not published, rather than disappearing from the page.
- API model ID
- Not published
- Context window
- 1.05M
- Maximum output
- Not sourced yet
- Knowledge cutoff
- Not sourced yet
- Input modalities
- Not sourced yet
- Output modalities
- Not sourced yet
- Parameters
- Not sourced yet
- Availability
- Not sourced yet
- Cloud regions
- Not tracked yet
- Lifecycle
- Current
- API capabilities
- Tool calling, structured outputs, and batch support are not tracked yet
- Prompt caching
- Published at $0.20 per million cached input tokens
- Self-host
- Weight availability is pending
- Rate limits
- Not tracked yet
How to read this profile
The visual layer above carries the decisions. These notes preserve the model, ranking, coverage, and family context behind the numbers.
Kimi K3.1 Pro ranks #13 of 216 on the public leaderboard with a score of 74.61/100. It does not yet have enough sourced coverage for a verified position.
Kimi K3.1 Pro has no verified publication yet for weight access. Its 1.05M context window is documented separately from that pending weight status. The profile records its reasoning mode as reasoning.
Its explicit predecessor is Kimi K3. 47 of 782 tracked benchmark slots currently have displayable evidence. Missing categories stay blank.
Its strongest eligible category is Multimodal & Grounded at #1, while its lowest eligible position is Knowledge at #13. particularly strong for screenshots, documents, charts, and grounded multimodal workflows.
Last updated October 29, 2026. Runtime fields remain blank until a sourced snapshot exists.
Questions
How does Kimi K3.1 Pro perform overall in AI benchmarks?
Kimi K3.1 Pro ranks #13 out of 216 models on the public BenchAlign leaderboard, with a score of 74.61/100. Its evidence status is Estimated, and this profile shows 47 source-displayable benchmark rows. The label describes evidence depth, not a provider quality claim; inspect category rows before choosing a workload.
Is Kimi K3.1 Pro good for knowledge and understanding?
Kimi K3.1 Pro ranks #13 out of 179 eligible models for knowledge and understanding, with a public category score of 75.1/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Kimi K3.1 Pro good for coding and programming?
Kimi K3.1 Pro has source-displayable benchmark coverage for coding and programming, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is Kimi K3.1 Pro good for agentic tool use and computer tasks?
Kimi K3.1 Pro ranks #9 out of 135 eligible models for agentic tool use and computer tasks, with a public category score of 73.2/100. That places it in the current top ten for this category. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Kimi K3.1 Pro good for multimodal and grounded tasks?
Kimi K3.1 Pro ranks #1 out of 58 eligible models for multimodal and grounded tasks, with a public category score of 100/100. That places it in the current top ten for this category. Check the underlying rows before treating the aggregate as a workload guarantee.
Does Kimi K3.1 Pro have full benchmark coverage on BenchLM?
No. Kimi K3.1 Pro currently has 72 source-displayable rows. The profile exposes published, non-generated evidence and leaves missing categories blank until an exact evaluation is available. Coverage describes how much was measured; it is not a penalty added to an individual benchmark result.
What is the context window size of Kimi K3.1 Pro?
Kimi K3.1 Pro has a reported context window of 1.05M in the exact-model catalog record. The value stays visible, but the profile marks its source link as unavailable instead of presenting it as directly documented. Maximum output length remains separate because providers often publish a different limit.