Skip to content
OSMEXO
Sign inGet an API key
☰
Sign inGet an API key

Kimi K3.1 Pro

moonshot/kimi-k3.1-pro
  • Moonshot AI
  • Family · Kimi K3 Pro
  • Released 23 Oct 2026
  • Text · Image · Video → Text
  • Reasoning · adjustable
  • Tools · JSON · Vision
Not on OSMEXO yet
Request this model
Spec & list price
Context1M
Max output944K
Input / 1M$2.00
Output / 1M$10.005× input
Scores · standing
Intelligence55.1#6/489
LiveBench71.3#10/32
Usage
Tokens / day571.6B
Usage rank#10
Requests / day1.9M
Tool calls / req0.71
Per request · 30 days
Avg input115.7K7.8× median
Avg output1.1K1.6× median
Cache hit94%1.4× median
Reasoning share62%about median
Tool-call failures0.1%0.0× median

Consumption

200.0B400.0B600.0B26 Oct29 Oct1 Nov
Peak 525.8B 30 OctLow 82.2B 23 OctRank today #10

News & trends

385 mentions this week ▲ 702%this model · videos, posts, articles · 30 days
Trending · 14 days
News · latest from Moonshot AI

Benchmarks

bar = share of the top score · Benchmark data: Artificial Analysis
Terminal-Bench 4.0#1 of 152
67.2%
Long-context reasoning#3 of 393
86.3%
Intelligence index#6 of 489
55.1
SciCode#9 of 160
60.8%
Humanity's Last Exam#11 of 465
52.5%
Speed · median across 2 providers, latest day
Output speed#277/33438 tok/s1.7× slower than median
Fastest provider53 tok/sbest of 2 providers
First token#3175.38 s5.2× longer than median
First answer token#9473 sthinking included

Best value in its score band

Value
Intelligence Index
5254565860$2$5$10Kimi K3.1 Pro
Blended price per 1M tokens (log)
LiveBench suites
Browser75.9
Coding81.8
Data Analysis79.1
Design77.8
Long Agentic39.1
Reasoning91.1
Research77.9

▲ Strongest: Reasoning, Coding. ▼ Weakest: Long Agentic (39).

Technical details

Reasoning
on · max
Tokenizer
Other
Moderation
none
Reasoning effort
  • max · default
  • high
  • low
  • none
Request parameters 13
  • tools
  • tool_choice
  • response_format
  • structured_outputs
  • reasoning
  • reasoning_effort
  • include_reasoning
  • max_tokens
  • stop
  • frequency_penalty
+3 more
  • logprobs
  • presence_penalty
  • top_logprobs

Interesting facts

  1. 7.4× the output room of similar models (943.7K against a 128.0K median).
  2. Released 10 days ago.

When to use

Best for

  • Coding agents on the hardest tasks
  • Long codebase and document reasoning
  • Video and image understanding
  • Scientific and research-grade reasoning
  • Very large single-response outputs

Think twice for

  • Long autonomous agent runs
  • Self-hosting (closed weights)
  • High-volume simple chat (cheaper options exist)

Alternatives

other vendors · similar capability

Want Kimi K3.1 Pro on OSMEXO?

Sign in to request it. Requests are reviewed; a request does not guarantee availability.

Request this model on OSMEXO