MoonshotAI: Kimi K3
MoonshotAI: Kimi K3 is moonshotai's model. On the un-leaked meo benchmark it ranks #13 of 82 with 80.3% accuracy and an Effective Value (π) of 0.1535 β earning a correct answer for about $0.0379 in 32s each.
as of 2026-10-18
Per-domain accuracy
202 correct of 250 questions asked across 11 domains Β· answer rate 100% Β· macro accuracy 89.7%. Headline accuracy is micro-averaged per meoβs methodology; per-domain question counts (correct/asked) are shown so you can judge significance.
Efficiency
- Cost per correct
- $0.0379
- Seconds per correct
- 32
- Tokens per correct
- 2,532
- Correct answers
- 202 / 250
- Total benchmark cost
- $7.66
- Avg tokens / item
- 2,046
Model details
- Lab
- Moonshot AI
- Released
- 2026-07-16
- Weights
- Open
- Context
- 1.0M
- Max output
- 944K
- Knowledge cutoff
- β
- Price in
- $0.45/M
- Price out
- $15.00/M
- Modality
- text+image+video->text
Sources: openrouter Β· meo
FAQ
How accurate is MoonshotAI: Kimi K3?
MoonshotAI: Kimi K3 scores 80.3% overall on the meo benchmark (rank #13 of 82), an un-leaked private holdout graded objective-first.
What is MoonshotAI: Kimi K3's Effective Value?
Its Effective Value (π) is 0.1535 at N=10, Ο=1, Ξ΄=1.5 β a fused metric of accuracy, speed, cost, and error-cascade. π rankings shift with task depth.
How much does MoonshotAI: Kimi K3 cost per correct answer?
About $0.0379 per correct answer (2,532 tokens, 32 seconds each) at benchmark-time pricing via OpenRouter.
Is MoonshotAI: Kimi K3 open weights?
Yes β MoonshotAI: Kimi K3 is released with open weights.
Benchmark updates affecting this model
- 2026-10-15majorAdded Gemini 4 Argon, Claude Opus 5.5, GPT-6.1 Sol, Kimi K3 and Grok 4.7
First board refresh since June: five models released JulyβOctober 2026 benchmarked over the full 250-item holdout under the unchanged standard protocol (reasoning effort high, temperature 1, separate answer budget, FINAL-line answer protocol, via OpenRouter). All five are vision-capable; each first passed a vision check that required describing a rendered holdout image. 1,250 responses, 0 errors, 2 empty replies (scored 0). Battery cost: Grok 4.7 $10.36 (longest reasoning, ~8.6k output tokens per text item), Kimi K3 $7.66, Gemini 4 Argon $5.68, Claude Opus 5.5 $3.79, GPT-6.1 Sol $0.82.
- 2026-06-09majorlogic_math_cs ground truth was broken β culled and rebuilt with solver-verified items
A 5-lab frontier jury revealed that 29 of 37 LLM-authored logic_math_cs items had broken ground truth: on most, five independent frontier models unanimously concluded the clue sets were contradictory while the stored answer claimed a number, and on eight, all five agreed on the SAME different number (author arithmetic errors). Models were being marked wrong for being right. The 29 were retired (full audit published in the repo) and the domain rebuilt with generator-as-oracle items: an embedded constraint solver PROVES each puzzle has exactly one solution (or provably none), verified by an independent exhaustive enumerator. A 3-round difficulty study found frontier models saturate consistent human-scale logic puzzles, so items use anonymous global constraints and carry a measured difficulty gradient.
- 2026-06-08majorCorrected base_bias grading (LLM-equivalence judge fix)
The LLM-equivalence grader uses a reasoning model whose verdict token budget was too small, so its YES/NO answer was emitted as empty and silently counted as a failure. This understated the base_bias domain to ~7% for every model when the true accuracy is ~85-90%. We raised the judge budget, excluded its reasoning tokens from the verdict, switched to verbosity-robust conclusion matching, and re-graded the whole board.
Embed this ranking
Show MoonshotAI: Kimi K3βs meo rank on your site β the badge links back here.
Related: all Moonshot AI models Β· Unsatisfiable constraints domain Β· best Effective Value Β· biggest movers Β· methodology Β· full leaderboard.