Specs & substance
- Developed by
- Moonshot
- Model family
- Kimi
- Use case
- Multimodal agentic
- Modality
- Text + Vision
- Context window
- 1M tokens
- Architecture
- MoE + KDA
- Version
- K3
- License
- Kimi K3 License
- Pricing
- $3.00 in · $15.00 out · $0.30 cache read
- Released
- Jul 2026
- Endpoint
- parasail-kimi-k3
Kimi K3 is a frontier open-weight model for long-horizon autonomy, combining reasoning with native text+image understanding in one model. It is designed for coding with minimal oversight across large repositories and terminal tool orchestration, as well as agentic knowledge work.
K3 has 2.8T total parameters with 104B activated. It is built on Kimi Delta Attention (KDA) and Attention Residuals (AttnRes) with a Stable LatentMoE framework activating 16 of 896 experts across 93 layers composed of 69 KDA + 24 Gated MLA. Its MoonViT-V2 vision encoder has 401M parameters. K3 uses MXFP4 weights with MXFP8 activations and was trained with quantization-aware training from the SFT stage onward, with a 1M-token context window.
Kimi K3 runs on Parasail via Elastic endpoints and Dedicated deployments, with OpenAI-compatible access and capacity sized to your workload.
Measured on the work that matters
Kimi K3 model-card results alongside selected comparison models.
All scores are self-reported by Moonshot in the Kimi K3 model card on Hugging Face ; the comparison columns are in turn cited there from Artificial Analysis, Vals AI, and each benchmark’s official leaderboard. Where a benchmark is reported both with and without tool augmentation, we show the lower, no-tools number.
Drop-in via the OpenAI SDK
Point any OpenAI-compatible client at Parasail and change the model name. That's it.
from openai import OpenAI
client = OpenAI(
api_key="YOUR_API_KEY",
base_url="https://api.parasail.io/v1"
)
response = client.chat.completions.create(
model="parasail-kimi-k3",
messages=[
{"role": "user", "content": "Hello, what can you do?"}
],
stream=True,
max_tokens=1000
)
for chunk in response:
if chunk.choices and chunk.choices[0].delta.content is not None:
print(chunk.choices[0].delta.content, end="", flush=True)curl https://api.parasail.io/v1/chat/completions \
-H "Authorization: Bearer $PARASAIL_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "parasail-kimi-k3",
"messages": [{"role": "user", "content": "Hello, what can you do?"}],
"stream": true,
"max_tokens": 1000
}'Explore the library
Kimi K2.7 Code
Specialized coding model optimized for software engineering and agentic coding workflows.
Kimi K2.6
Strong reasoning model with excellent long-context understanding.
GLM-5.2
Z.AI’s next-gen model for agentic engineering with sustained execution on long-horizon tasks.
GLM-5.3
Z.AI’s strongest GLM-5 agentic coder — same base as 5.2, all the gains from post-training, open-source SOTA on coding and cyber benchmarks.
Bring Kimi K3 to your workload.
Request access and tell us what you're building — we'll help get you set up.