Models / Kimi K3
Kimi K3
Kimi K3 is best for writing, and reasoning and knowledge; pick it when quality matters more than cost. The cheapest model of similar intelligence is Claude Haiku 5.5, at $0.20 per million tokens against $6.00.
Results 16 Nov 2026 · 101 results · 11 sources
Test Kimi K3 on your prompt → It runs beside the models you could use instead, with every answer checked and costed.
- Price per million tokens
- $3.00 in · $15.00 out
Moonshot AI's list price, 10 Nov 2026 · compare prices$0.30 cached input - Speed
- 41 tokens a second; on its test prompt the answer starts after 51.8 s, thinking time included
Artificial Analysis, on its usual API, at max reasoning; 40–41 across 2 settings - Intelligence Index
- 43.6
Artificial Analysis, at max reasoning; 30.1–43.6 across 2 settings - Spring Prompt overall
- 19th of 36 score 47 of 100 · how it works
- Spring Prompt Score
- 37 of 100 23rd of 32 · 4 of 5 benchmarks · our own tests only
- Developer
- Moonshot AI, based in China
Weights published - Released
- 16 Jul 2026
1,048,576 tokens of context
Where it stands among the models you could choose
Available models grouped by price, intelligence, speed and developer location, plus models with published weights. Retired models are excluded.
| Among | Intelligence Index | Spring Prompt overall | Price, blended | Speed |
|---|---|---|---|---|
| Models available today | 19th of 185Claude Opus 5.5 | 19th of 35Claude Fable 5.2 | $6.00 blended, 90th percentilecheapest: Mistral Nemo | 63rd of 65Nemotron 3.5 Super VL |
| Models priced $3–10 per million tokens | 11th of 38Claude Opus 5.5 | 11th of 15Claude Opus 5.5 | sets this group | 13th of 15Mistral Medium 3.5 |
| Models with similar intelligence | sets this group | 3rd of 9GPT-6 Sol | $6.00 blended, 83rd percentilecheapest: Claude Haiku 5.5 | 10th of 12Claude Haiku 5.5 |
| Models writing under 50 tokens a second | 3rd of 9MiMo-V2.6-Pro | – | $6.00 blended, 100th percentilecheapest: Phi-4 | sets this group |
| Models from developers based in China | 5th of 75MiMo-V2.6-Pro | 4th of 11Kimi K3.1 Pro | $6.00 blended, 100th percentilecheapest: Qwen3.5-9B | 21st of 23DeepSeek-V4.1-Flash |
| Open-weights models | 4th of 104MiMo-V2.6-Pro | 3rd of 8Kimi K3.1 Pro | $6.00 blended, 100th percentilecheapest: Mistral Nemo | 42nd of 43Nemotron 3.5 Super VL |
The name under each rank is the leader of that group; green is the top quarter of the group and red the bottom quarter, by rank. Price is shown as a percentile: the share of the group that costs less, so lower is cheaper. "Sets this group" marks the measure the group is defined by. Spring Prompt overall is our score out of 100 across every source (how it works). Price is the developer's list price (or the typical OpenRouter host where we haven't read one) for three input tokens to one output; intelligence and speed are from Artificial Analysis (data sourced from Artificial Analysis); "similar intelligence" means within 4 points on its Intelligence Index. Developer locations are where each company is based, not where a model is served. Groups under 3 models are not ranked.
Measured cost per task
API cost recorded in our benchmark runs, excluding judging.
| Task | Benchmark | Cost per task |
|---|---|---|
| Product listing | CatalogBench | ≈ $0.047 |
| Deck | DeckBench | ≈ $0.454 |
| Game of bullet chess | BulletBench | ≈ $0.054 |
Cheapest model that passes
On our benchmarks that define a pass for each task: what one passing answer cost in our run, failed attempts included, and whether a cheaper model passed.
- CatalogBench · cost per 1,000 accepted listings
Kimi K3: $76.49 (103 of 168 attempts), 23 of 25. Cheapest: GPT-6 Luna at $0.70 (114 of 168 attempts). It passed at least as often. - DeckBench · cost per presentable deck
Kimi K3: no presentable decks (0 of 6 tasks). Cheapest: GPT-6.1 Terra at $0.13 (3 of 6 tasks).
Cost per attempt divided by the share that passed, at the run date's prices, excluding judging. Each model at its cheapest setting per passing answer. With a handful of tasks, a pass or two either way is within chance. No estimate where a run had missing outputs, because their cost may not have been recorded.
How good is it, and for what?
Each use case ranks the available models its benchmarks measured, then those priced $3–10 per million tokens. The bar shows its percentile in that field, best to the right. Open a row for the results behind it.
Product listingsTurning a sparse product feed and photos into listings that can go live
| Benchmark | Kimi K3 | Rank among available | Best available |
|---|---|---|---|
| CatalogBenchCatalogBench: reliably publish-ready | 41.1% | 17th of 32 | GPT-6 Astra 75.0% |
| CatalogBenchCatalogBench (sales brief): reliably publish-ready | 1.8% | 25th of 32 | GPT-6 Astra 69.6% |
Decks from an analysisTurning a finished analysis into a deck you could present as it is
| Benchmark | Kimi K3 | Rank among available | Best available |
|---|---|---|---|
| DeckBenchDeckBench: deck rating | 1,205 | 14th of 32 | Qwen4-Max 1,585 |
User surveysPlanning a user survey and reading its results without being misled
- SurveyBench: has not measured this model.
Marketing planningPlanning a year of ad spend without overspending
| Benchmark | Kimi K3 | Rank among available | Best available |
|---|---|---|---|
| ROASBenchROASBench: overall score | 44.1 | 17th of 32 | Claude Fable 5.2 61.3 |
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls
| Benchmark | Kimi K3 | Rank among available | Best available |
|---|---|---|---|
| Artificial AnalysisArtificial Analysis: Terminal-Bench 4.0 · reported | 12.6% | 35th of 107 | Kimi K3.1 Pro 67.2% |
| Artificial AnalysisArtificial Analysis: Terminal-Bench 2.1 · reported | 85.0% | 11th of 107 | Claude Fable 5.1 91.4% |
| Artificial AnalysisArtificial Analysis: τ-bench banking · reported | 46.0% | 9th of 104 | Grok 4.6 50.7% |
| Vending-Bench 2Vending-Bench 2: money after a year · reported | $5,165 | 23rd of 53 | GPT-6 Astra $15,515 |
- tau2-bench: pending. Its latest results are dated 26 May 2026, before this model was released.
- Berkeley Function Calling Leaderboard (BFCL) V4: pending. Its latest results are dated 16 Dec 2025, before this model was released.
- OpenHands Index: pending. Its latest results are dated 30 Jun 2026, before this model was released.
- Microsoft STATE-Bench: pending. Its latest results are dated 29 May 2026, before this model was released.
Professional workReal tasks from banking, consulting and law, business documents and freelance projects
| Benchmark | Kimi K3 | Rank among available | Best available |
|---|---|---|---|
| APEX-AgentsAPEX-Agents: tasks passed · reported | 50.6% | 21st of 37 | Claude Sonnet 5.5 75.5% |
| GDP.pdfGDP.pdf: professional document tasks · reported | 19.0% | 17th of 31 | GPT-5.6 Sol 30.7% |
- Remote Labor Index: has not measured this model.
Reasoning and knowledgeHard questions across science, maths and general knowledge
| Benchmark | Kimi K3 | Rank among available | Best available |
|---|---|---|---|
| Artificial AnalysisArtificial Analysis: Artificial Analysis Intelligence Index · reported | 43.6 | 19th of 185 | Claude Opus 5.5 57.6 |
| Artificial AnalysisArtificial Analysis: Humanity's Last Exam · reported | 46.9% | 20th of 186 | Claude Opus 5.5 61.4% |
| Artificial AnalysisArtificial Analysis: GPQA Diamond · reported | 93.5% | 10th of 168 | GPT-6 Astra 96.3% |
WritingWhat people prefer in blind comparisons, and judged writing quality
| Benchmark | Kimi K3 | Rank among available | Best available |
|---|---|---|---|
| Arena (formerly LMArena)Arena Text: overall · reported | 1,488 | 12th of 153 | Gemini 4 Argon 1,525 |
- UGI Leaderboard: has not measured this model.
Sticking to the factsSummarising without inventing things, and factual answers
| Benchmark | Kimi K3 | Rank among available | Best available |
|---|---|---|---|
| Arena (formerly LMArena)Arena Text factuality: overall · reported | 1,472 | 17th of 109 | Gemini 4 Argon 1,511 |
| SimpleQA Verified (Epoch AI)SimpleQA Verified: correct answers · reported | 50.6% | 22nd of 68 | GPT-6 Astra 75.6% |
- Vectara Hallucination Leaderboard: has not measured this model.
Speed under pressureGood decisions against a real clock (fast chess)
| Benchmark | Kimi K3 | Rank among available | Best available |
|---|---|---|---|
| BulletBenchBulletBench Bullet 60s: ladder Elo | 33 | 30th of 33 | Gemini 3.8 Flash 1,068 |
Nearest alternatives
- Cheaper at similar intelligenceClaude Haiku 5.5$0.20 blended · Intelligence Index 43.4Compare with Kimi K3 →
Against Kimi K3 at $6.00 blended, Intelligence Index 43.6. "Similar intelligence" means within 4 points or better.
Where to run it
24 hosts on OpenRouter at the standard service tier, with a median of $5.14 blended per million tokens; quantisation where the host states it: FP4, MXFP4, FP8. Cheapest first.
| Host | Quantisation | Input | Output | Blended |
|---|---|---|---|---|
| Relace | FP4 | $0.18 | $9.34 | $2.47 |
| Wafer | MXFP4 | $0.28 | $9.60 | $2.61 |
| Sail Research | FP4 | $0.28 | $13.26 | $3.52 |
| Sail Research | FP4 | $0.28 | $13.26 | $3.52 |
| ParetoAlpha | not stated | $1.00 | $11.25 | $3.56 |
| Chutes | MXFP4 | $1.85 | $9.50 | $3.76 |
| Decart | MXFP4 | $1.95 | $9.75 | $3.90 |
| InferenceNet | FP4 | $0.23 | $15.00 | $3.93 |
Other service tiers, not counted above: Wafer fast $0.15 in, $16.00 out; InferenceNet fast $0.45 in, $18.00 out; Parasail fast $4.50 in, $22.50 out; Fireworks fast $4.50 in, $22.50 out. Flex and batch are cheaper but slower or queued; priority and fast cost more.
Prices per million tokens from OpenRouter's public endpoint list, checked 10 Nov 2026; the 16 dearer hosts are left out.
Against the alternatives
The models you would most likely weigh it against: the leaders of the groups above.
Scroll sideways to see every alternative.
| Kimi K3 | Claude Opus 5.5 most intelligent at $3–10 per million tokens | Claude Haiku 5.5 cheapest with similar intelligence | MiMo-V2.6-Pro most intelligent from China | Claude Fable 5.2 highest-scoring overall available today | |
|---|---|---|---|---|---|
| Price per million tokens | $6.00 | $8.00 | $0.20 | $0.54 | $16.00 |
| Tokens a second | 41 | 96 | 240 | 43 | – |
| Intelligence Index | 43.6 | 57.6 | 43.4 | 46.3 | – |
| Spring Prompt overall | 47 | 82 | 40 | – | 90 |
| Rank among available models, by use case | |||||
| Product listings | 21st of 32 | 13th | 20th | – | 5th |
| Decks from an analysis | 14th of 32 | 7th | 18th | – | 4th |
| User surveys | – | 4th | 7th | – | 1st |
| Marketing planning | 17th of 32 | 9th | 30th | – | 1st |
| Agents and tool use | 40th of 159 | 3rd | 28th | 23rd | – |
| Professional work | 24th of 51 | 3rd | – | – | – |
| Reasoning and knowledge | 18th of 187 | 1st | 25th | 13th | – |
| Writing | 12th of 169 | 5th | – | 23rd | – |
| Sticking to the facts | 27th of 139 | 3rd | – | 10th | – |
| Speed under pressure | 33rd of 35 | – | 21st | – | 8th |
| Side by side → | Side by side → | Side by side → | Side by side → | ||
Shaded figures are better than Kimi K3's. A dash means no result.
Research and examples
- Kimi K3.1 Pro first look: the best open-weight model on our board, and very slow
- We Told 11 AI Models What Happened in 2026. Most of Them Thought We Were Making It Up.
- The Anatomy Skill for Claude Code: A Practical Guide, Tested on Everyday Objects and Seven Models
In the playground: our pick in 1 of 20 tests, 96% of checks passed →
- A pricing table with a monthly/yearly switch (our pick)
- Holographic trading cards (5 of 5 checks)
- A map of the world from memory (4 of 4 checks)
Where it has been measured
3rd highest-scoring overall of the 9 models with similar intelligence, and 3rd most intelligent of the 9 models writing under 50 tokens a second. 11th most intelligent of the 38 models priced $3–10 per million tokens. Among models available today, in the top quarter (by percentile) for 4 of the 9 use cases measured, led by writing, and reasoning and knowledge; in the bottom quarter for speed under pressure.
- BulletBench5 results
- CatalogBench31 results
- DeckBench17 results
- ROASBench9 results
- APEX-Agents1 result
- Arena (formerly LMArena)11 results
- Artificial Analysis22 results
- GDP.pdf1 result
- SimpleQA Verified (Epoch AI)1 result
- tau2-bench2 results
- Vending-Bench 21 result
- Berkeley Function Calling Leaderboard (BFCL) V4Pending: latest results 16 Dec 2025
- Microsoft STATE-BenchPending: latest results 1 Jul 2026
- OpenHands IndexPending: latest results 30 Jun 2026
- SurveyBenchNot measured
- Remote Labor IndexNot measured
- UGI LeaderboardNot measured
- Vectara Hallucination LeaderboardNot measured
Every result
101 published results from 11 sources, each in the source's own units, with the configuration that produced it.
BulletBench · measured by Spring Prompt · 5 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| BulletBench Bullet 60s: cost per gameUS dollars, lower is better | $0.0540 | 62 of 71 | Kimi K3 (low reasoning) low reasoning |
1 Oct 2026 |
| BulletBench Bullet 60s: games lost on time% of games, lower is better | 91.7% | 68 of 71 | Kimi K3 (low reasoning) low reasoning |
1 Oct 2026 |
| BulletBench Bullet 60s: invalid moves% of moves, lower is better | 0.0% | 1 of 71 | Kimi K3 (low reasoning) low reasoning |
1 Oct 2026 |
| BulletBench Bullet 60s: ladder Eloladder Elo, higher is better | 33 0–270 |
66 of 71 | Kimi K3 (low reasoning) low reasoning |
1 Oct 2026 |
| BulletBench Bullet 60s: median move timemilliseconds, lower is better | 2.6 s | 53 of 71 | Kimi K3 (low reasoning) low reasoning |
1 Oct 2026 |
Not shown: Not your cost: prices are those charged on the run date.
CatalogBench · measured by Spring Prompt · 31 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| CatalogBench (sales brief): channel rules broken% of products, lower is better | 0.6% | 17 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): claims to check% of products, lower is better | 0.0% | 1 of 32 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): content quality% of checks, higher is better | 96.4% | 20 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): failed outputs% of products, lower is better | 0.6% | 26 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): missing UK information% of products, lower is better | 0.0% | 1 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): not findable% of products, lower is better | 3.0% | 14 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): publish-ready listings% of products, higher is better | 14.9% | 24 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): reliably publish-ready% of products, higher is better | 1.8% | 25 of 34 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): unsupported claims% of products, lower is better | 81.6% | 29 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): unsupported claimsclaims per product, lower is better | 3.64 | 28 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): wrong attributes% of products, lower is better | 20.8% | 25 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): wrong category or variant% of products, lower is better | 4.2% | 29 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: channel compliance% of products, higher is better | 99.4% | 26 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: channel rules broken% of products, lower is better | 0.6% | 20 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: claims to check% of products, lower is better | 0.0% | 1 of 32 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: conflicts caught% of conflicts, higher is better | 97.0% | 33 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: content quality% of checks, higher is better | 96.4% | 23 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: cost per productUS dollars, lower is better | $0.0469 | 33 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: decision accuracy% of decisions, higher is better | 96.1% | 20 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: failed outputs% of products, lower is better | 0.0% | 1 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: field accuracy% of missing fields, higher is better | 94.4% | 8 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: invented values% of filled values, lower is better | 5.7% | 34 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: missing UK information% of products, lower is better | 0.0% | 1 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: not findable% of products, lower is better | 4.2% | 17 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: publish-ready listings% of products, higher is better | 61.3% | 18 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: reliably publish-ready% of products, higher is better | 41.1% | 17 of 34 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: time per productseconds, lower is better | 66.6 | 32 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: unsupported claims% of products, lower is better | 16.7% | 17 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: unsupported claimsclaims per product, lower is better | 0.98 | 25 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: wrong attributes% of products, lower is better | 24.4% | 31 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
| CatalogBench: wrong category or variant% of products, lower is better | 3.0% | 24 of 35 | Kimi K3 provider default reasoning |
30 Sep 2026 |
Not shown: Not your catalogue: invented products with generated images, so a model's result on your own feed can differ.
DeckBench · measured by Spring Prompt · 17 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| DeckBench: accurate decks% of tasks, higher is better | 50.0% | 15 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: caveat dropped% of tasks, lower is better | 0.0% | 1 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: clean layout% of tasks, higher is better | 33.3% | 16 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: cost per deckUS dollars, lower is better | $0.45 | 30 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: deck ratingrating, higher is better | 1,205 1,150–1,292 |
14 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: design quality% of the maximum, higher is better | 81.4% | 5 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: draft figure quoted% of tasks, lower is better | 0.0% | 1 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: findings missing% of tasks, lower is better | 16.7% | 23 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: head-to-head win rate% of comparisons, higher is better | 69.8% | 11 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: layout defects% of tasks, lower is better | 66.7% | 16 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: misleading metric used% of tasks, lower is better | 0.0% | 1 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: presentable decks% of tasks, higher is better | 0.0% | 15 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: recommendation late or wrong% of tasks, lower is better | 33.3% | 20 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: slides needing work% of tasks, lower is better | 16.7% | 3 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: time per deckseconds, lower is better | 438 | 31 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: unsupported claims% of tasks, lower is better | 33.3% | 24 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
| DeckBench: unsupported numbers% of tasks, lower is better | 0.0% | 1 of 34 | Kimi K3 provider default reasoning |
6 Oct 2026 |
Not shown: Not your analysis or your brand: invented companies, six public tasks, one deck per model per task at the provider's default reasoning setting.
ROASBench · measured by Spring Prompt · 9 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| ROASBench: audience scorescore out of 100, higher is better | 65.1 | 15 of 35 | Kimi K3 provider default reasoning |
29 Sep 2026 |
| ROASBench: business scorescore out of 100, higher is better | 40.2 | 17 of 35 | Kimi K3 provider default reasoning |
29 Sep 2026 |
| ROASBench: consistency scorescore out of 100, higher is better | 28.3 | 18 of 35 | Kimi K3 provider default reasoning |
29 Sep 2026 |
| ROASBench: contribution profitsimulated US dollars, higher is better | $654,270 | 15 of 35 | Kimi K3 provider default reasoning |
29 Sep 2026 |
| ROASBench: cost of a runUS dollars, lower is better | $1.01 | 30 of 35 | Kimi K3 provider default reasoning |
29 Sep 2026 |
| ROASBench: months over budgetmonths of 12, lower is better | 0 | 1 of 35 | Kimi K3 provider default reasoning |
29 Sep 2026 |
| ROASBench: overall scorescore out of 100, higher is better | 44.1 | 17 of 35 | Kimi K3 provider default reasoning |
29 Sep 2026 |
| ROASBench: planning scorescore out of 100, higher is better | 54.7 | 20 of 35 | Kimi K3 provider default reasoning |
29 Sep 2026 |
| ROASBench: return on ad spendprofit per $1 spent, higher is better | $0.50 | 19 of 35 | Kimi K3 provider default reasoning |
29 Sep 2026 |
Not shown: Not real ad performance: the market is a deterministic simulation of one invented skincare brand.
APEX-Agents · reported by APEX-Agents · 1 result
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| APEX-Agents: tasks passed% of tasks, higher is better | 50.6% | 22 of 39 | Kimi K3 | 1 Oct 2026 |
Not shown: Not your firm's documents or tools; graded by rubric, not by a client.
Arena (formerly LMArena) · reported by Arena (formerly LMArena) · 11 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| Arena Agent: confirmed task successIPS effect estimate, higher is better | 0.08 0.06–0.09 |
9 of 50 | Kimi K3 (max reasoning, arena agent) | 2 Oct 2026 |
| Arena Agent: praise over complaintIPS effect estimate, higher is better | 0.07 0.05–0.09 |
17 of 50 | Kimi K3 (max reasoning, arena agent) | 2 Oct 2026 |
| Arena Agent: steerabilityIPS effect estimate, higher is better | 0.01 0.01–0.02 |
24 of 50 | Kimi K3 (max reasoning, arena agent) | 2 Oct 2026 |
| Arena Agent: tool groundingIPS effect estimate, higher is better | 0.00 0.00–0.01 |
2 of 50 | Kimi K3 (max reasoning, arena agent) | 2 Oct 2026 |
| Arena Text factuality: overallArena rating, higher is better | 1,472 1,469–1,476 |
25 of 182 | Kimi K3 (max reasoning) | 2 Oct 2026 |
| Arena Text: business, management and financeArena rating, higher is better | 1,487 1,479–1,496 |
14 of 406 | Kimi K3 (max reasoning) | 2 Oct 2026 |
| Arena Text: creative writingArena rating, higher is better | 1,460 1,451–1,469 |
26 of 411 | Kimi K3 (max reasoning) | 2 Oct 2026 |
| Arena Text: expert promptsArena rating, higher is better | 1,533 1,521–1,545 |
11 of 363 | Kimi K3 (max reasoning) | 2 Oct 2026 |
| Arena Text: instruction followingArena rating, higher is better | 1,488 1,481–1,495 |
14 of 413 | Kimi K3 (max reasoning) | 2 Oct 2026 |
| Arena Text: overallArena rating, higher is better | 1,488 1,484–1,493 |
16 of 413 | Kimi K3 (max reasoning) | 2 Oct 2026 |
| Arena Text: writing, literature and languageArena rating, higher is better | 1,476 1,468–1,484 |
16 of 412 | Kimi K3 (max reasoning) | 2 Oct 2026 |
Not shown: Not a success rate: the estimated effect of choosing this model as the agent's orchestrator, against an even mix of the models tested, inside Arena's own harness.
Artificial Analysis · reported by Artificial Analysis · 22 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| Artificial Analysis: Artificial Analysis Coding Indexindex score, higher is better | 72.0 | 39 of 187 | Kimi K3 (low reasoning) | 20 Oct 2026 |
| Artificial Analysis: Artificial Analysis Coding Indexindex score, higher is better | 76.2 | 20 of 187 | Kimi K3 (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Artificial Analysis Intelligence Indexindex score, higher is better | 30.1 | 115 of 413 | Kimi K3 (low reasoning) | 20 Oct 2026 |
| Artificial Analysis: Artificial Analysis Intelligence Indexindex score, higher is better | 43.6 | 43 of 413 | Kimi K3 (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: GPQA Diamond% of questions, higher is better | 84.2% | 134 of 360 | Kimi K3 (low reasoning) | 20 Oct 2026 |
| Artificial Analysis: GPQA Diamond% of questions, higher is better | 93.5% | 14 of 360 | Kimi K3 (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Humanity's Last Exam% of questions, higher is better | 25.0% | 179 of 419 | Kimi K3 (low reasoning) | 20 Oct 2026 |
| Artificial Analysis: Humanity's Last Exam% of questions, higher is better | 46.9% | 42 of 419 | Kimi K3 (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Long-context reasoning (AA-LCR)% of questions, higher is better | 79.3% | 96 of 407 | Kimi K3 (low reasoning) | 20 Oct 2026 |
| Artificial Analysis: Long-context reasoning (AA-LCR)% of questions, higher is better | 88.7% | 1 of 407 | Kimi K3 (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Output speedtokens per second, higher is better | 39.9 | 129 of 137 | Kimi K3 (low reasoning) | 20 Oct 2026 |
| Artificial Analysis: Output speedtokens per second, higher is better | 41.0 | 128 of 137 | Kimi K3 (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: SciCode% of problems, higher is better | 52.7% | 83 of 205 | Kimi K3 (low reasoning) | 20 Oct 2026 |
| Artificial Analysis: SciCode% of problems, higher is better | 59.5% | 14 of 205 | Kimi K3 (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Terminal-Bench 2.1% of tasks, higher is better | 82.4% | 38 of 188 | Kimi K3 (low reasoning) | 20 Oct 2026 |
| Artificial Analysis: Terminal-Bench 2.1% of tasks, higher is better | 85.0% | 26 of 188 | Kimi K3 (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Terminal-Bench 4.0% of tasks, higher is better | 12.6% | 86 of 204 | Kimi K3 (low reasoning) | 20 Oct 2026 |
| Artificial Analysis: Terminal-Bench 4.0% of tasks, higher is better | 12.6% | 86 of 204 | Kimi K3 (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Time to first answer tokenseconds, lower is better | 53.0 | 120 of 137 | Kimi K3 (low reasoning) | 20 Oct 2026 |
| Artificial Analysis: Time to first answer tokenseconds, lower is better | 51.8 | 118 of 137 | Kimi K3 (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: τ-bench banking% of tasks, higher is better | 41.6% | 26 of 175 | Kimi K3 (low reasoning) | 20 Oct 2026 |
| Artificial Analysis: τ-bench banking% of tasks, higher is better | 46.0% | 12 of 175 | Kimi K3 (max reasoning) | 20 Oct 2026 |
Not shown: Not your codebase or tools.
GDP.pdf · reported by GDP.pdf · 1 result
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| GDP.pdf: professional document tasks% of rubric, higher is better | 19.0% | 25 of 49 | Kimi K3 (max reasoning) | 1 Oct 2026 |
Not shown: Not your documents; 100 tasks, so small differences are noise.
SimpleQA Verified (Epoch AI) · reported by SimpleQA Verified (Epoch AI) · 1 result
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| SimpleQA Verified: correct answers% of questions, higher is better | 50.6% | 23 of 85 | Kimi K3 (max reasoning) | 27 Aug 2026 |
Not shown: Not answers grounded in your documents; tests what the model remembers.
tau2-bench · reported by tau2-bench · 2 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| tau2 v1.0.1, banking knowledge, alltools retrieval: consistency (pass^4)% of tasks, higher is better | 17.5% | 12 of 21 | Kimi K3 (max reasoning, tau2) | 24 Jul 2026 |
| tau2 v1.0.1, banking knowledge, alltools retrieval: task success (pass^1)% of tasks, higher is better | 37.1% | 11 of 21 | Kimi K3 (max reasoning, tau2) | 24 Jul 2026 |
Not shown: Not your policies or customers: the tasks and policies are tau2's own, and the customer is simulated by gpt-5.2.
Vending-Bench 2 · reported by Vending-Bench 2 · 1 result
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| Vending-Bench 2: money after a yearUS dollars, higher is better | $5,165 | 24 of 63 | Kimi K3 | 1 Oct 2026 |
Not shown: Not a real business; one simulated market with set rules.
Embed a score badge
A badge for its Spring Prompt Score and each of our benchmarks that ranked Kimi K3, from the current results. Paste the code into a page or README; it links back to where the result comes from.
Spring Prompt Score · 4 of 5 benchmarks
<a href="https://springprompt.com/score#moonshotai-kimi-k3"><img src="https://springprompt.com/badge/score/moonshotai-kimi-k3.svg" alt="Kimi K3 is #23 of 32 on the Spring Prompt Score (37 of 100, from 4 of 5 of our benchmarks), measured by Spring Prompt" height="20"></a>
CatalogBench · Reliably publish-ready, factual brief
<a href="https://springprompt.com/benchmarks/catalogbench?focus=moonshotai-kimi-k3"><img src="https://springprompt.com/badge/catalogbench/moonshotai-kimi-k3.svg" alt="Kimi K3 is #17 of 32 on CatalogBench (Reliably publish-ready, factual brief: 41.1%), measured by Spring Prompt" height="20"></a>
DeckBench · Deck rating
<a href="https://springprompt.com/benchmarks/deckbench?focus=moonshotai-kimi-k3"><img src="https://springprompt.com/badge/deckbench/moonshotai-kimi-k3.svg" alt="Kimi K3 is #14 of 33 on DeckBench (Deck rating: 1,205), measured by Spring Prompt" height="20"></a>
ROASBench · Score
<a href="https://springprompt.com/benchmarks/roasbench?focus=moonshotai-kimi-k3"><img src="https://springprompt.com/badge/roasbench/moonshotai-kimi-k3.svg" alt="Kimi K3 is #17 of 33 on ROASBench (Score: 44.1), measured by Spring Prompt" height="20"></a>
BulletBench · Ladder Elo
<a href="https://springprompt.com/benchmarks/bulletbench?focus=moonshotai-kimi-k3"><img src="https://springprompt.com/badge/bulletbench/moonshotai-kimi-k3.svg" alt="Kimi K3 is #56 of 59 on BulletBench (Ladder Elo: 33), measured by Spring Prompt" height="20"></a>
Will Kimi K3 work for your product?
Benchmarks aren't your task. Paste the prompt your product uses and run it beside the models you could use instead, with every answer checked and costed.
New here? Your email starts a free trial. No credit card needed.