Kimi K3
Moonshot AI · text, image, video → text
Kimi K3 is Kimi’s flagship model for long-horizon coding and end-to-end knowledge work, with a 1M-token context window and industry-leading intelligence.
Input$3.00 /M
Output$15.00 /M
← ModelsKimi K3
Qwen3 VL 235B A22B Thinking
Kimi K3 vs Qwen3 VL 235B A22B Thinking
Compare Kimi K3 from Moonshot AI and Qwen3 VL 235B A22B Thinking from Qwen on key metrics including benchmarks, price, context length, and other model features. Access both models and hundreds of others through the AIHubMix API.
Kimi K3Qwen3 VL 235B A22B Thinking+ Add model · 2/4
Qwen3 VL 235B A22B Thinking
Qwen · text, image, video → text
The Qwen3 series open-source models include hybrid models, thinking models, and non-thinking models, with both reasoning capabilities and general abilities reaching industry SOTA levels at the same scale.
Input$0.27 /M
Output$2.74 /M
Pricing & Specifications
Prices are per million tokens. Time to First Token and throughput are rolling averages measured on AIHubMix.
Kimi K3
Qwen3 VL 235B A22B Thinking
Input /M
$3.00
$0.27
Output /M
$15.00
$2.74
Cache read /M
$0.30
$0.27
Context length
1,048,576
131,000
Max output
1,048,576
33,000
Time to First Token
0.9 s
2.2 s
Throughput
73.1 tok/s
57.8 tok/s
Modalities
textimagevideo
textimagevideo
Supported Parameters
thinkingfunction callingstructured outputs
thinkingtoolsfunction callingstructured outputs
API Formats
Released
-
-
Promotional prices show the discounted rate; see each model page for promotion windows.
Activity Past 30 Days
Daily traffic served through AIHubMix — how demand for each model is trending.
kimi-k3qwen3-vl-235b-a22b-thinking
Tokens / day
-
Requests / day
-
Performance Past 3 Days
Measured on real AIHubMix traffic, hourly buckets. Gaps mean no traffic in that hour.
kimi-k3qwen3-vl-235b-a22b-thinking
Throughput (tok/s)
-
TTFT (s)
-
Uptime (%)
-
LMArena Benchmarks
LMArena ratings by capability (Bradley-Terry, commonly called Elo). Higher is better.
Text
kimi-k3qwen3-vl-235b-a22b-thinking
13201380144015001560
Overall
13961486
Coding
14561531
Math
14051405
Hard prompts
14191505
Instruction following
13841479
Multi-turn
13881499
Creative writing
13391470
Longer query
14071502
Chinese
14401527
English
14041491
Source: LMArena (arena.ai) leaderboard, imported by AIHubMix. Models without published ratings are omitted per chart.
Cost calculator
Estimate your monthly bill for the same workload on each model.
Qwen3 VL 235B A22B Thinking
$57.54 /mo
Kimi K3
$405 /mo
Monthly = daily × 30. Discounted rates applied where a promotion is active.
FAQ
Which is cheaper: Kimi K3, Qwen3 VL 235B A22B Thinking?
Qwen3 VL 235B A22B Thinking: $2.74/M output tokens; Kimi K3: $15.00/M. Use the cost calculator above to estimate your own workload.
How do their coding arena scores compare?
Kimi K3: 1531; Qwen3 VL 235B A22B Thinking: 1456 (LMArena coding leaderboard).
Which responds faster?
Kimi K3: 0.9s time to first token measured on AIHubMix; see the live performance charts above for how each model behaves across the day.
How large is each context window?
Kimi K3 accepts 1,048,576 and Qwen3 VL 235B A22B Thinking accepts 131,000 input tokens. Maximum output per request is 1,048,576 tokens on Kimi K3 and 33,000 tokens on Qwen3 VL 235B A22B Thinking.
Which one generates tokens faster?
Kimi K3 at 73.1 tok/s and Qwen3 VL 235B A22B Thinking at 57.8 tok/s, measured as output throughput on AIHubMix — a separate metric from time to first token.
What inputs and capabilities does each model support?
Kimi K3 accepts text, image and video input and supports thinking, function calling and structured outputs; Qwen3 VL 235B A22B Thinking accepts text, image and video input and supports thinking, tool calling, function calling and structured outputs.
Can I call Kimi K3 and Qwen3 VL 235B A22B Thinking with the same API key?
Yes. AIHubMix serves every model on this page behind one OpenAI-compatible endpoint, so switching between them is a one-line change to the model field — no second account, key or SDK.
Popular comparisons
Related model match-ups readers also look at.
AIHubMix