Kimi K3

moonshotai/kimi-k3-20260715

MoonshotAI Kimi K3, from MoonshotAI, accepts text and image inputs and produces text output. It offers a context length of 1,048,576 tokens and supports tool usage and reasoning capabilities. Pricing is set at 3 units per million input tokens and 15 units per million output tokens.

  • tool use
  • structured outputs
  • reasoning
  • open-weight

At a glance

providermoonshotai
releasedJul 16, 2026
context1.0M
input / M$0.24

Specifications

What the API gives you

Context window
1.0M tokens
Max output
944K tokens
Input price
$0.24 / 1M tokens
Output price
$11.94 / 1M tokens
Cache read
$0.21 / 1M tokens
Modalities
text, image, video → text
Knowledge cutoff
—
Released
Jul 16, 2026

Measured evidence

Public benchmark record

Design Arena Coding
1362 Elo2026-10-20
Design Arena Full-stack Agent
1315 Elo2026-10-14
Design Arena UI
1346 Elo2026-10-20
Design Arena Web-app Agent
1292 Elo2026-10-20
Design Arena Website
1336 Elo2026-10-20
FrontierMath Tiers 1-3 (Epoch AI)‡
72.2% ± 2.72026-07-17max effort, evaluated 2026-07-17
GPQA Diamond (Epoch AI)‡
91.9% ± 1.92026-08-07high effort, evaluated 2026-08-07
GPQA Diamond (Epoch AI)‡
93.1% ± 1.52026-07-16max effort, evaluated 2026-07-16
LiveBench Agentic Coding
62.2%2026-10-18
LiveBench Coding
81.5%2026-10-18
LiveBench Data Analysis
78.7%2026-10-18
LiveBench Instruction Following
71.4%2026-10-18
LiveBench Language
85.5%2026-10-18
LiveBench Math
84.4%2026-10-18
LiveBench Reasoning
90.7%2026-10-18
LMArena Agent‡
1.92026-10-19max effort
LMArena Agent Bash Recovery‡
4.42026-10-08max effort
LMArena Agent Steerability‡
0.52026-10-19max effort
LMArena Agent Task Outcome‡
6.42026-10-19max effort
LMArena Agent Tool Hallucination‡
0.72026-10-19max effort
LMArena Creative Writing‡
1456 Elo2026-10-15max effort
LMArena Instruction Following‡
1483 Elo2026-10-15max effort
LMArena Long Query‡
1493 Elo2026-10-15max effort
LMArena Math‡
1494 Elo2026-10-15max effort
LMArena Text‡
1487 Elo2026-10-15max effort
LMArena Text Factuality‡
1471 Elo2026-10-15max effort
LMArena WebDev‡
1655 Elo2026-10-19max effort
LMArena WebDev Image-to-Web‡
1579 Elo2026-10-19max effort
OTIS Mock AIME (Epoch AI)‡
93.3% ± 3.82026-08-07high effort, evaluated 2026-08-07
OTIS Mock AIME (Epoch AI)‡
97.2% ± 1.12026-07-16max effort, evaluated 2026-07-16
SimpleQA Verified (Epoch AI)‡
50.6% ± 1.62026-08-27max effort, evaluated 2026-08-27
tau2-bench (1.0.1)‡
37.1%2026-07-24max effort, evaluated 2026-07-24

LMArena data licensed CC-BY-4.0. Aider (Apache-2.0) and SWE-bench (MIT) are open. Design Arena uses blind human preference. ‡ Submitted scaffold plus model result. Compare the harness and effort settings, not only the model name.

Open-weight model on Hugging Face.

Task evidence

How broad is the support?

Each status is exact to this model version and the selected task. Official release claims remain provisional.

  • Best LLM for Competition MathEstablished evidence

    100% of intended weighted task evidence across 4 independent sources. 6 of 6 configured evaluation feeds currently have results.

  • Best LLM for Business WritingEstablished evidence

    100% of intended weighted task evidence across 2 independent sources. 4 of 4 configured evaluation feeds currently have results.

  • Best LLM for Logic & Problem SolvingEstablished evidence

    100% of intended weighted task evidence across 3 independent sources. 5 of 5 configured evaluation feeds currently have results.

  • Best LLM for Marketing CopyEstablished evidence

    100% of intended weighted task evidence across 2 independent sources. 4 of 4 configured evaluation feeds currently have results.

  • Best LLM for Chat & RoleplayEstablished evidence

    100% of intended weighted task evidence across 2 independent sources. 4 of 4 configured evaluation feeds currently have results.

  • Best LLM for Creative WritingEstablished evidence

    100% of intended weighted task evidence across 2 independent sources. 4 of 4 configured evaluation feeds currently have results.

Show all 41 tasks
  • Best LLM for Frontend & UIEstablished evidence

    100% of intended weighted task evidence across 3 independent sources. 6 of 6 configured evaluation feeds currently have results.

  • Best LLM for WritingEstablished evidence

    100% of intended weighted task evidence across 2 independent sources. 4 of 4 configured evaluation feeds currently have results.

  • Best LLM for Roleplay & CharacterEstablished evidence

    100% of intended weighted task evidence across 2 independent sources. 4 of 4 configured evaluation feeds currently have results.

  • Best LLM for Math & ReasoningEstablished evidence

    100% of intended weighted task evidence across 5 independent sources. 7 of 7 configured evaluation feeds currently have results.

  • Best LLM for a General AssistantEstablished evidence

    100% of intended weighted task evidence across 3 independent sources. 6 of 6 configured evaluation feeds currently have results.

  • Best LLM for Code CompletionEstablished evidence

    79% of intended weighted task evidence across 2 independent sources. 3 of 3 configured evaluation feeds currently have results.

  • Best Open-Weight LLM for Self-HostingEstablished evidence

    64% of intended weighted task evidence across 2 independent sources. 2 of 3 configured evaluation feeds currently have results.

  • Best LLM for Tool & Function CallingEstablished evidence

    58% of intended weighted task evidence across 2 independent sources. 6 of 7 configured evaluation feeds currently have results.

  • Best LLM for Fast SummariesEstablished evidence

    58% of intended weighted task evidence across 2 independent sources. 2 of 3 configured evaluation feeds currently have results.

  • Best Value LLMEstablished evidence

    58% of intended weighted task evidence across 2 independent sources. 2 of 3 configured evaluation feeds currently have results.

  • Best LLM for Agentic CodingEstablished evidence

    55% of intended weighted task evidence across 3 independent sources. 7 of 7 configured evaluation feeds currently have results.

  • Best LLM for CodingEstablished evidence

    53% of intended weighted task evidence across 3 independent sources. 7 of 7 configured evaluation feeds currently have results.

  • Best LLM for Literary TranslationEstablished evidence

    50% of intended weighted task evidence across 2 independent sources. 3 of 4 configured evaluation feeds currently have results.

  • Best LLM for Everyday TranslationEstablished evidence

    50% of intended weighted task evidence across 2 independent sources. 3 of 4 configured evaluation feeds currently have results.

  • Best LLM for TranslationEstablished evidence

    50% of intended weighted task evidence across 2 independent sources. 3 of 4 configured evaluation feeds currently have results.

  • Best Local LLM for CodingEstablished evidence

    47% of intended weighted task evidence across 2 independent sources. 4 of 4 configured evaluation feeds currently have results.

  • Best LLM for AI AgentsEstablished evidence

    43% of intended weighted task evidence across 3 independent sources. 9 of 9 configured evaluation feeds currently have results.

  • Best LLM for Document SummarizationEstablished evidence

    40% of intended weighted task evidence across 2 independent sources. 5 of 6 configured evaluation feeds currently have results.

  • Best LLM for SummarizationEstablished evidence

    40% of intended weighted task evidence across 2 independent sources. 5 of 6 configured evaluation feeds currently have results.

  • Best Cheap LLM for High VolumeEstablished evidence

    40% of intended weighted task evidence across 2 independent sources. 2 of 3 configured evaluation feeds currently have results.

  • Cheapest Capable LLMEstablished evidence

    40% of intended weighted task evidence across 2 independent sources. 2 of 3 configured evaluation feeds currently have results.

  • Best LLM for Autonomous AgentsEstablished evidence

    36% of intended weighted task evidence across 2 independent sources. 6 of 6 configured evaluation feeds currently have results.

  • Best LLM for JSON & Schema OutputEstablished evidence

    35% of intended weighted task evidence across 2 independent sources. 3 of 4 configured evaluation feeds currently have results.

  • Best LLM for Structured Data ExtractionEstablished evidence

    35% of intended weighted task evidence across 2 independent sources. 3 of 4 configured evaluation feeds currently have results.

  • Best LLM for Debugging & Code ReviewEstablished evidence

    33% of intended weighted task evidence across 2 independent sources. 5 of 5 configured evaluation feeds currently have results.

  • Best LLM for Grounded RAGEstablished evidence

    29% of intended weighted task evidence across 2 independent sources. 5 of 7 configured evaluation feeds currently have results.

  • Best LLM for Hermes AgentProvisional evidence

    29% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or SWE-rebench. 4 of 4 configured evaluation feeds currently have results.

  • Best LLM for OpenHandsProvisional evidence

    29% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or SWE-rebench. 4 of 4 configured evaluation feeds currently have results.

  • Best LLM for GooseProvisional evidence

    29% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or SWE-rebench. 4 of 4 configured evaluation feeds currently have results.

  • Best LLM for OpenClawProvisional evidence

    29% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or SWE-rebench. 4 of 4 configured evaluation feeds currently have results.

  • Best LLM for RAG & Long ContextProvisional evidence

    22% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from FACTS Grounding or LMArena Document. 5 of 6 configured evaluation feeds currently have results.

  • Best LLM for High-Volume ExtractionProvisional evidence

    21% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Route reliability or Structured-output eval. 1 of 3 configured evaluation feeds currently have results.

  • Best LLM for Massive ContextProvisional evidence

    20% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Document or LongBench v2. 2 of 3 configured evaluation feeds currently have results.

  • Best LLM for Computer & Browser UseProvisional evidence

    17% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or OSWorld Verified. 3 of 3 configured evaluation feeds currently have results.

  • Best LLM for Image UnderstandingProvisional evidence

    11% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Vision or VLMEvalKit tasks. 2 of 3 configured evaluation feeds currently have results.

Recommendation coverage

Where this model appears

  1. #1Best Local LLM for Coding
  2. #3Best LLM for Structured Data Extraction
  3. #3Best LLM for JSON & Schema Output
  4. #4Best Open-Weight LLM for Self-Hosting
  5. #4Best LLM for Code Completion
  6. #5Cheapest Capable LLM
  7. #5Best LLM for High-Volume Extraction
  8. #5Best Cheap LLM for High Volume
  9. #6Best LLM for Massive Context
  10. #6Best LLM for Debugging & Code Review
  11. #6Best Value LLM
  12. #6Best LLM for Agentic Coding
  13. #7Best LLM for Autonomous Agents
  14. #7Best LLM for AI Agents
  15. #8Best LLM for Fast Summaries

Compare head to head

Catalog data via OpenRouter and models.dev. Last checked within the past 6 hours.